Orthogeriatric co-management improves the outcome of long-term care residents with fragility fractures

Fragility fractures are a major health care problem worldwide. Due to increasing life expectancy and other associated demographic changes, the incidence of fractures and post-fracture disability appear certain to increase [1]. Fragility fractures are typically caused by osteoporosis, primarily affecting postmenopausal women, but also older men. Four out of every 10 white women age 50 or older in the United States will experience a hip, spine, or wrist fracture sometime during their lives, while 13 % of white men will suffer a similar fate [2]. Both hip fractures and non-hip fractures are associated with excess mortality in the years following the fracture [3].

The increased mortality risk lasts for 5–10 years post-fracture but is most pronounced in the first 3–6 months after sustaining a hip fracture [4, 5]. The reasons for this increased mortality risk are poorly understood [4]. Despite advances in surgical and medical care, the excess mortality of hip fracture patient remains high and has not improved over the last decade [6]. Excess mortality after hip fracture may be linked to complications following the fracture, such as pulmonary embolism, infections, and heart failure. Risk factors associated with falls and additional osteoporotic fractures may contribute to high mortality rates [7, 8]. Individual characteristics of persons sustaining a hip fracture likely play an important role, e.g., low-bone density is associated with increased non-trauma mortality, even without fractures [9]. Poor functional status is also independently linked to poor outcomes; impairments in daily activities and low-mobility scores are associated with a higher mortality [10].

Residents of long-term nursing homes represent a high-risk group for both mortality and poor functional outcomes. They are twice as likely to sustain hip fractures, and their post-fracture outcomes are worse than among community dwellers [11‐13]. Many studies on hip fracture patients exclude nursing home residents or are limited by small sample size, single center design, and lack of data on functional outcomes. There is little published data looking at this cohort in particular and the impact of non-hip fragility fractures. While orthogeriatric co-management models of care have been shown in multiple studies to have medical as well as economic advantages [14, 15], there are little data to assess the impact of co-management on the long-term outcomes of this high-risk group of residents. Our study is the first to focus on the outcomes of long-term care residents after a fragility fracture (hip and non-hip) initially treated under an orthogeriatric co-management model.

Our data set contained 256 long-term care residents who were hospitalized for the treatment of a fragility fracture. The majority of our patients were female (80 %). 130 (49.1 %) sustained a hip fracture, and 135 (50.9 %) sustained a non-hip fracture. All patient’s characteristics are shown in Table 1. As expected, hip fracture patients underwent significantly more frequent surgery and had a longer length of stay than non-hip fracture patients. Except for the CAM score, we found no significant differences between the two fracture groups. This cohort had a high number of comorbidities and low level of pre-fracture functionality. The most frequent comorbidities were cerebrovascular diseases (60 %), dementia (53.6 %), heart failure (52.5 %), polypharmacy (50.2 %), sarcopenia (38.5 %), urinary incontinence (29.1 %), depression (27.9 %), recurrent falls (27.5 %), malnutrition (24.2 %), hearing impairment (23.8 %), diabetes (22.7 %), atrial fibrillation (22.3 %), heart valve disease (20.8 %), chronic pain syndrome (16.6 %), and renal failure (15.5 %). The prevalence of all other diseases and geriatric syndromes was below 15 %. Mobility was severely impaired in the majority of our patients. 164 (62.6 %) had a pre-fracture PMS of lower than four points, 25 (9.4 %) were immobile with a PMS of 0, and only 15 (5.7 %) had no mobility deficit. Only one out of five were discharged to a rehabilitation unit (20.9 %). Patients admitted to a rehabilitation unit were significantly younger (85.3 ± 6.4 vs. 87.2 ± 6.4; p = 0.038), and had a better pre-fracture mobility (4.4 ± 2.6 vs. 3.2 ± 2.0; p = 0.001) and a lower CAM score (0.6 ± 0.9 vs. 1.2 ± 1.3; p = 0.007).

Following admission, the lowest BI was noted on post-operative day 5 (mean 31.2 ± 21.0) with hip fracture patients demonstrating a lower BI (26.2 ± 20.3) compared to non-hip fracture patients (38.6 ± 19.9). In both fracture groups, we observed an improvement in independence from day 5 to 3 month follow-ups (overall BI improvement 17.2 points, hip fracture 19.1, and non-hip fracture 14.4). At the 6 and 12 month follow-ups, we found a negligible improvement in the BI.

187 (70.6 %) patients died during the observation time. The mean life expectancy after the index fracture was 527 (±431) days. Males had a non-significantly shorter survival time compared to female patients (458 ± 403 days vs. 546 ± 438 days; p = 0.322). Within the first 12 months, 78 (29.4 %) patients died, with no statistically significant differences based on fracture group. Age, low BMI, CCI, SGS, PMS, and CAM were significantly associated with 1-year mortality (Table 2). After logistic regression, including sex, age, fracture site, surgical vs. conservative treatment, LOS, BMI, CCI, SGS, PMS, and CAM, only older age (p = 0.013; OR 1.091; CI 1.019–1.169) and a lower PMS (p = 0.022; OR 0.737; CI 0.568–0.957) remained as independent predictors of one-year mortality. Figure 2 shows the one-year mortality risk in different age groups by different PMS groups.

During follow-up, 62 (23.4 %) patients sustained one subsequent fracture and 10 (3.4 %) patients sustained more than one fracture; overall, 29 (10.9 %) patients had recurrent fractures within the first year (Fig. 1). Fracture group, surgical treatment, CCI, SGS, PMS, and CAM were significantly associated with the presence of a subsequent fracture (Table 3). After logistic regression, including sex, age, fracture group, operative vs. non-operative treatment, LOS, BMI, CCI, SGS, PMS, and CAM, none of them remained as independent predictor of the subsequent fracture.

The poor outcome of the long-term care residents after a fragility fracture in usual care settings is already well known [10, 22, 23]. P. Orthogeriatric co-management models have shown beneficial effects on the outcomes of older fragility fracture patients [14, 15]. Data regarding special subgroups, such as long-term care residents, are rare. Our study is the first to focus on the long-term-care residents treated under an orthogeriatric co-management model. A mean CCI was 3.3, SGS of more than 6 and PMS of 3.3 reflect the high number of comorbidities and disabilities in our study population compared to other fragility fracture cohorts [10, 24]. 38.5 % of our patients had a CCI of 4 or higher. In a comparison cohort, Neuman et al. found a lower percentage of patients with similar CCI scores (26.6 %) [25]. As we expected, life expectancy is extremely limited in these patients. The mean survival time in our study was relatively long at 527 days as compared to the cohort studied by Neuman et al. at 377 days [25]. Compared to the results of Neuman, who described a 1-year mortality in hip fracture patients of 47 %, we found a much lower mortality rate [25]. One-year mortality in our study was 29.4 % in the overall group and 33.8 % in hip fracture patients. For hip fracture patients, we calculated a significantly reduced one-year mortality in patients treated in our orthogeriatric co-management model with an OR of 0.57 (95 % CI 0.31–0.85). Poor outcomes among long-term nursing home residents with hip fractures have previously been noted in other small trials. Berry et al. noted a mortality rate of 40 % at 1 year, Beaupre et al. noted 45 %, and Morrsion and Siu noted a 6-month mortality of 55 % [26‐28]. Other than our orthogeriatric co-management model, we did not find any other reason for the lower mortality rate. However, we are not able to describe, how our model leads to a decrease of mortality. Our orthogeriatric co-management model is based on multifactorial, interdisciplinary interventions.

Among our study, population's higher age and a lower PMS were the only independent risk factors for one-year mortality. Surprisingly, CCI and others factors did not remain significant predictors after adjustment for other risk factors. Older patients with a low PMS had the highest one-year mortality risk. Figure 2 shows the one-year mortality risk in different age groups and PMS groups (9–7, 6–4, 3–0 points).

Secondary fracture prevention is a main goal of orthogeriatric co-management. 23.4 % of all patients sustained a subsequent fracture during observation time, 10.9 % within the first year. Even in the group of survivors, the rate of subsequent fractures remained low at 26.9 %. The follow-up time for this cohort was long at more than 4 years (mean 1417 days, 95 % CI 954–1842 days). For comparison, we calculated 573 person years for the overall cohort. From our data, we calculated a risk of any subsequent fracture at 108 per 1000 person years. This was comparable to the result of Center et al., who assessed similar risks 90 per 1000 person years in community-dwelling men and women over the age of 80 [29]. In light of similar outcomes, despite significant differences in fracture risk, we postulate that our orthogeriatric approach likely reduces the risk of subsequent fractures. The quantity of this benefit cannot be determined without more comparable study populations or a more rigorous study design. In this study, fracture risk approximates a J-curve with the lowest rate of the subsequent fractures (20.7 %) in patients with a PMS lower than 4 and the highest in patients with a score from 7 to 9 (37 %). In long-term care patients, immobility may reduce the risk of subsequent fractures, and we found higher rates of fracture in patients with higher PMS. This could have an impact on decisions regarding osteoporosis treatment. In our study, the risk of a subsequent fracture was much lower than the mortality risk and older long-term care patients with low PMS may not benefit from osteoporosis treatment.

Our trial is the first to evaluate both hip and non-hip fracture patients admitted from the long-term care settings to geriatric fracture center. Despite the overall high mortality and poor post-operative mobility, orthogeriatric co-management seems to be able to improve the outcome of high-risk fragility fracture patients. One-year mortality was 29.4 % and significant lower than that found in comparable trials. Orthogeriatric co-management also seems to have an impact on functional outcome and the risk of subsequent fractures. Further trials and analysis are necessary to quantify the benefit of co-management for this special group of patients, particularly in regard to how it influences the outcome.