Impact of orthogeriatric management on the average length of stay of patients aged over seventy five years admitted to hospital after hip fractures

This was an observational, retrospective cohort single-centre study conducted at our Orthopaedic Trauma Unit between January 1, 2017, and December 31, 2018. Patients aged ≥ 75 with an X-ray diagnosis of hip fracture requiring surgical repair were eligible for inclusion. The exclusion criterion was the presence of a concomitant or pathological fracture.

In January 2018, an orthogeriatric unit, supervised by a geriatrician, has been set up within our orthopaedic department. This orthogeriatric unit allows a systematic pre-operative evaluation of all patients over 75 years of age suffering from a hip fracture. This unit also deals with post-operative medical care.

Patients operated on by January 2018 were able to benefit from orthogeriatric care unlike patients operated on between January 1, 2017, and December 31, 2018. Thus, the patients were divided into group 1 (no orthogeriatric care), and group 2 (received orthogeriatric care).

The following patient data were obtained:

The comorbidities identified can be categorized as cardiovascular (hypertension, dyslipidaemia and diabetes), vascular (obliterative arterial disease of the lower limbs, transient ischaemic attack or stroke, angina and coronary syndrome), degenerative (dementia and Parkinson’s disease), cardiac (arrhythmia and heart failure), pulmonary (asthma and chronic obstructive pulmonary disease) or psychiatric. Other comorbidities include prior history of cancer, kidney failure (moderate or severe), tobacco use and alcoholism.

The Charlson Comorbidity Index score, which predicts the risk of mortality, was calculated for each patient [11].

We compared groups 1 and 2 in terms of average length of stay, preoperative delay, intrahospital mortality, discharge rate, mortality after 6 months, and the number of new fractures that occurred within six months of the first fracture.

Categorical data are presented as numbers and proportions, and the groups were compared using chi-squared tests or file tests when the theoretical numbers were too low (n < 5). Quantitative data are presented as mean and standard deviation, and the groups were compared using Student’s t test. Statistical analyses were conducted using SAS software (ver. 9.4 for Windows, SAS Institute, Inc., Cary, NC, USA).

We included a total of 534 patients in this study. The two groups were comparable in terms of epidemiological characteristics and comorbidities (Tables 1 and 2). In total, 32 patients (6%) were lost to follow-up during the study (16 patients in each group; 6.5% and 5.5% of groups 1 and 2, respectively; p = 0.85).

The two groups were similar in terms of fracture types and treatments (Table 3). There was a statistically significant difference between the groups only in the rate of compression plate fixation (p = 0.02). More patients were treated with compression plate fixation in 2017 than in 2018. General anaesthesia was performed preferentially in 398 cases (77%). Spinal anaesthesia was performed in 119 cases (23%). There was a significant difference in the rate of spinal anaesthesia between the two groups (p < 0.0001), which appeared unrelated to the physiological status of the patients, instead reflecting different preferences regarding anaesthesia type (Table 3). The two groups were similar in terms of the rates of the various complications (Table 4).

Table 4 presents the average length of stay, postoperative recovery time and discharge rate data. There was a statically significant difference between the groups only in average length of stay, which was 1.8 days shorter in Group 2 (95% CI: 0.64; 2.59; p = 0.001).

The six month mortality rate was 22.5%. There was no significant group difference in the six month mortality rate (58 [23.6%] and 61 [21.2%] patients in groups 1 and 2, respectively; p = 0.64). Figures 1 and 2 show the six month mortality rate (based on the Charlson score) for both groups; 90% of patients with a Charlson score ≤ 5 were alive after six months, and 50% of patients with a Charlson score ≥ 10 had died after 6 months.

Eleven new fractures occurred in group 1, compared to eight in group 2, at six months post-operatively (p = 0.75) (Table 4).

Orthogeriatric unit treatment reduced the median length of stay by one day, in line with most previous studies (Table 5). For example, in the metaanalysis of Kammerlander et al., the average length of stay across was 22.7 and 26.6 days in orthogeriatric treatment and control groups, respectively [12]. However, no such decrease was reported in other studies [13, 14]. According to Pablos-Hernandez et al. [2], multifaceted orthogeriatric treatment is the most effective (orthogeriatric unit treatment plus daily home visits by a geriatrician), and could reduce the surgical treatment time and thus the length of stay. Prestmo et al. found that joint management of patients by an orthopaedic surgeon and geriatrician improved the prognosis compared to orthopaedic management involving only the mobile geriatric unit [15]. In our study, only 38% of patients received treatment within 48 hours, while 65% had surgery beyond 48 hours after being admitted to hospital. In the study by Doshi et al., 37% of patients were operated on within 48 hours [16], compared to 87% of patients in the study of Boddaert et al. [17]. Timely treatment is not only dependent on orthogeriatric unit treatment but also on the availability of operating theatres and therapeutics, particularly anticoagulants. In our study, 22% of patients had a cardiac arrhythmia and 18% were treated with anticoagulants. Almost half of the patients on anticoagulants had a pre-operative delay of more than 72 hours, comparable to the rate reported by Boddaert et al. [17]. Anticoagulant treatment alone cannot explain the delays. Several studies showed that pre-operative delays were associated with an increase in post-operative complications such as confusion [18], pressure ulcers [18] and lung infections [19]. It is clear that hip fracture is an emergency condition, and a metaanalysis of 35 studies reported that elderly hip fracture patients operated on within the first 48 hours after admission had a significantly lower mortality rate than those with a delay of > 72 h [20]. It also seems important to use a scoring system to determine patients requiring surgical management within 24 hours. Such systems are instructive for all medical personnel involved in patient care (i.e. surgeons, anaesthetists and geriatricians). Indeed, in 2003, a team in Belfast (Northern Ireland) proposed a scoring system to identify patients requiring treatment within the first 24 hours [21].

In our study, the intra-hospital mortality was 2.6%, consistent with existing literature; in a prospective cohort study by Doshi et al. including 219 patients, the intra-hospital mortality rate was 2.3% [16]. Orthogeriatric care did not affect the intra-hospital mortality rate in our study, consistent with a metaanalysis [22]. Although a decrease in the intra-hospital mortality rate after orthogeriatric care was reported by Friedman et al. (from 2.5 to 1.6%), it was not statistically significant [23]. However, Vidam et al. reported a significant reduction, from 5.8 to 0.6% [13].

The discharge rate was not different between the patients receiving and not receiving orthogeriatric care, consistent with previous studies [24, 25].

In our study, the six month mortality rate had decreased from 23.6 to 21.2% with the establishment of the orthogeriatric unit. In a six year study, Boddaert et al. reported six month mortality rates of 24% and 15% in their control groups and orthogeriatric treatment, respectively [17].

Orthogeriatric treatment did not reduce the rate of new fractures at six months post-operatively. Similarly, Gregersen et al. reported a nonsignificant decrease in the rate a new fractures after two years following hip fracture, with versus without orthogeriatric treatment (9.5% vs. 7.7%).

Limitations include a lack of generalizability due to the single-centre design and retrospective nature of this study. A retrospective cohort study depends on data found in the medical file such as looking for comorbidity and complications. However, we have limited the impact of this bias by using data that does not require clinical interpretation. Moreover, this retrospective study did not allow us to state the criteria for discharge of our patients. However, the only change made in our department between 2017 and 2018 is the establishment of an orthogeriatric unit. It seems therefore that orthogeriatric unit treatment is largely the cause of the observed differences. It is still possible that differences may be attributable to something other than the model of care like surgical approach, nursing care or a more effective social inclusion network. We also observed in our study the use of trochanteric nail and spinal anaesthesia more frequently in 2018 than in 2017. This appeared unrelated to the physiological status of the patients but seems to be related to changes in habits. However, it is quite possible that these observations have consequences on the post-operative cares leading to a reduction in the length of stay. In order to precisely answer these questions, we are currently conducting a prospective study to assess the influence of other data that may influence length of stay and six months mortality.