Impact of Delirium on Short-term Prognosis in Geriatric Patients with Hip Fracture

Background: To explore the effect of delirium on postoperative early prognosis of geriatric patients with hip fracture. Methods: According to whether postoperative delirium (POD) appeared in elderly patients with hip fracture, all cases were divided into two groups: delirium and non-delirium. General parameters, operative parameters, length of stay (LOS), and perioperative complications were recorded. Recurrent fracture, new stroke, and mortality within 1 year postoperatively were followed up and compared between the two groups. Results: Of 358 cases, 55 (15.4%) developed delirium. Compared with the non-delirium group, the delirium group had longer operative times (P=0.031); increased platelet (P=0.002) and C-reactive protein (P<0.001) levels; and reduced hemoglobin (P<0.001), calcium (P=0.040), albumin (P<0.001), prealbumin (P<0.001), and total cholesterol (P=0.019) levels. Complications occurred to some extent in 196 cases (54.7%). In the delirium group, complication rates of pulmonary infection (P=0.005), hypoalbuminemia (P<0.001), electrolyte disorder (P<0.001), dyspepsia (P=0.027), bedsore (P=0.012) and anemia (P=0.007) were higher. In addition, LOS was signicantly longer (15.71 ± 4.72 vs 14.00 ± 5.30 days; P=0.026) in the delirium group. Within 1 year postoperatively, 31 individuals (8.7%) had died, 13 were treated for recurrent fracture, and 42 had a new stroke. In the delirium group, mortality within 1 year postoperatively was higher (27.3% vs 5.3%, P<0.001). Conclusions: The morbidity of POD is high in geriatric patients with hip fractures. There was longer surgery delay,


Background
Hip fracture in elderly patients is one of the most common osteoporotic fractures. As society is increasingly aging, the amount of hip fractures is increasing 1  Postoperative delirium (POD) is a common clinical manifestation that occurs in 13-51% of elderly patients with hip fractures 4,5 . Several studies have demonstrated a correlation between POD and length of stay (LOS), as well as hospital costs [6][7][8] . Recently Hamilton et al. reported that POD could increase the postoperative mortality of patients who underwent non-cardiac surgery 9 . However, studies focusing on possible relationships between POD and short-term prognosis in geriatric patients with hip fractures are lacking.
In this study, we collected the clinical statistics of elderly patients with hip fractures treated in our hospital, with the aim of analyzing potential correlations between POD and short-term prognosis, such as LOS, early complications during the hospital stay, and postoperative mortality. December 2018. Among all cases, two resulted from car accidents, two occurred with pathological fractures, two occurred with old fractures, ve were treated with conservative treatments, and ten cases failed to follow up; all these cases were excluded according to the exclusion criteria. The remaining 358 cases were enrolled in this study, of which 121 patients were male and 237 patients were female. Ages ranged from 65 to 99 years, with an average of 79.67 ± 7.09 years. The time from injury to operation ranged between 1 and 20 days, with an average time of 6.86 ± 3.24 days. One hundred seventy-nine cases were injured on the left side, while 179 cases were injured on the right side. One hundred fortyseven cases occurred with femoral neck fracture, while 211 cases occurred with intertrochanteric fracture.
With regard to treatments, 223 cases underwent internal xation, 211 cases were xed with an intramedullary nail, 135 cases were treated with arthroplasty, and 12 cases adopted a cannulated screw.
Groups were divided according to whether POD occurred. According to the Confusion Assessment Method (CAM) described in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), all elderly patients with hip fracture were screened and evaluated 10 by a group of clinicians who had been trained . Assessment items included: (1) acute onset and uctuating course, (2) inattention, (3) disorganized thinking, and (4) altered level of consciousness. The diagnosis of delirium by CAM requires the presence of features 1, 2, and either 3 or 4. All patients were divided into two groups: delirium (55 cases) and nondelirium (303 cases).

Data collection and follow up methods
The following data of each case were recorded and collected. (1) General information and operative data such as gender, age, body mass index, preoperative coexisting diseases (e.g., hypertension, diabetes mellitus, chronic obstructive pulmonary disease, coronary heart disease, old myocardial infarction, old cerebral infarction, Parkinson's disease, hepatic dysfunction, and renal dysfunction), preoperative laboratory examination (e.g., blood routine, hepatic and renal function tests, coagulation test, in ammatory biomarkers), injury-to-operation time, American Society of Anesthesiologists (ASA) scale, type of anesthesia, type of operative, and operative time. (2) Length of stay: dates of the hospital stay. (3) Perioperative complications were checked and analyzed, such as pulmonary infection, deep venous thrombosis, urinary system infection, hypoproteinemia, electrolyte disorder, arthralgia, anchyloses, dyspepsia, bedsore, arrhythmia, cerebrovascular accident, multiorgan dysfunction, anemia. (4) All patients were followed up to record whether recurrent fracture, new stroke, or death occurred within 1 year by telephone or outpatient service after discharge.

Operative condition and perioperative management
The surgeons in charge of each patient had the quali cations of attending physician or above, and the capabilities of all doctors were comparable. All patients underwent routine examinations before treatment. Intraoperative spinal anesthesia was routinely used (only two patients underwent general anesthesia). Intertrochanteric fractures were treated with closed reduction and intramedullary nails. For non-displaced femoral neck fractures, hollow nails were used with internal xation; whereas, displaced femoral neck fractures were treated with hip replacement. Antibiotics were given 30 minutes before surgery and within 1 day after surgery to prevent infection. Non-steroidal analgesia and low molecularweight heparin anticoagulation were given during the perioperative period. Early functional training was started on the rst day after surgery. All patients were prescribed vitamin D and calcium for bone protection at discharge.

Statistical method
Data were collected and processed using MS Excel and SPSS 19.0 software. Measurement data are expressed as mean ± standard deviation. Comparisons between groups were performed using independent sample t-tests; counted data were analyzed by Chi-Square test. For non-normally distributed data, non-parametric tests were used. P < 0.05 was considered statistically signi cant.

Comparison of general data in delirium and non-delirium groups
A total of 358 older adult hip fracture cases were enrolled in this study, of which 55 had POD, yielding an incidence of 15.4%. As shown in Table 1 and Table 2, there was no signi cant difference in gender, age, body mass index, fracture type, xation method, number of comorbidities, ASA classi cation, or operation time between the two groups (P > 0.05). However, compared with the non-delirium group, the delirium group had a longer injury-to-operation time (P = 0.031); increased platelet counts (P = 0.002) and Creactive protein levels (P < 0.001); and reduced levels of serum calcium (P = 0.040), albumin (P < 0.001), prealbumin (P < 0.001), and total cholesterol (P = 0.019).   Table 3, the incidence of complications during hospitalization was high in both groups. Compared with the non-delirium group, the delirium group had a higher incidence of pulmonary infection (P = 0.005), hypoalbuminemia (P < 0.001), electrolyte disturbance (P < 0.001), indigestion (P = 0.027), bedsores (P = 0.012), and anemia (P = 0.007). In addition, hospital stays (15.71 ± 4.72 vs 14.00 ± 5.30 days; P = 0.026) of the delirium group were signi cantly longer compared with the non-delirium group.

Discussion
Compared with some research hotspots such as deep vein thrombosis, pulmonary embolism, and cardiovascular and cerebrovascular diseases of the lower extremities, POD is a common phenomenon, but it is easily ignored by orthopedic surgeons in elderly patients with hip fractures. Probably because POD is often an acute transient abnormality, resulting in missed diagnosis and misdiagnosis, which may lead to patients not receiving timely, reasonable, and effective intervention 6,[11][12][13] . In this study, we found that POD may be associated with a series of early adverse prognoses. Compared with the non-delirium group, the delirium group was more likely to have perioperative complications and the incidence of pulmonary infection, hypoalbuminemia, electrolyte disturbance, indigestion, bedsores, and anemia was higher. In addition, the length of stay was longer in the delirium group. Moreover, the mortality rate was 27.3% in the delirium group within one year after operation, which was ve times that of the non-delirium group.
Our results are similar to those of previous studies. Hamilton et al. 9 reviewed 34 of 4968 screened citations and found that delirium was associated with a four-fold increase in the risk of death. Mosk et al. 7 reported that delirium was correlated with a longer hospital stay, increased associated with complications, institutionalization, and 6-month mortality by observing 566 patients with hip fractures who were over the age of 70. Bellelli et al. 14 reported that hypoactivity and mixed POD were associated with increased 6-month mortality risk. They observed 571 patients in which the morbidity of delirium was 38.5%, whereas the mortality rates of non-delirium and delirium groups separately were 8.3% and 24.1%, respectively. Bai et al. 4 enrolled a total of 6288 patients and reported the prevalence of POD was 28% (23%-34%). They found that approximately one-fourth of patients undergoing hip fracture surgery went on to develop POD, and delirium increased both short-term and long-term mortality in these patients.
Recently, Bielza et al. 5 observed 383 patients with hip fracture who were over 70 years of age and found that 212 patients had POD; they found no correlation between delirium and mortality, walking ability, length of stay, or clinical complications. However, the in-hospital mortality of non-delirium and delirium groups was 4.6% and 3.3%, which is obviously lower than previously reported mortality rates of geriatric hip fractures. This difference may be related to the relatively short duration of their observation period, as they only collected data during the hospital stay. Therefore, there is a positive correlation between delirium and complications, as well as mortality, in our study and several similar studies. According to our results, early detection, early diagnosis, and early treatment of POD are very important and have positive clinical signi cance for reducing perioperative complications and the length of hospital stays, and improving mortality.  15 . Subsequent research on the mechanism of clinical POD caused by systemic in ammatory response has shown that acute peripheral in ammatory stimuli (such as infection, surgery, or trauma) can induce activation of parenchymal cells to promote the expression of proin ammatory cytokines (such as tumor necrosis factor, interleukin 6, and interleukin 1β) and chemokines (such as monocyte chemotactic protein 1) in the central nervous system, which causes destruction of the blood-brain barrier and subsequent neuron and synaptic dysfunction, resulting in delirium 16 . According to the results of this study, blood platelet counts and C-reactive protein levels were higher in patients with POD; whereas, hemoglobin, blood calcium, albumin, prealbumin, and total cholesterol levels were lower. These results indicate that delirium is indeed related to systematic in ammation, coagulation changes, and nutritional status of the whole body, not just local manifestations of the nervous system. In addition, Lee et al. 17 found that about 30.2% (70 cases) of older adult hip fracture patients had delirium after surgery, of which 20% (14 cases) continued to experience delirium 4 weeks after surgery, indicating that not all delirium is temporary. Indeed, it may sometimes appear to be continuous until the patient's whole-body condition is adjusted. Therefore, POD needs to be treated as soon as possible to be corrected in time. Previous studies reported that delirium results from the combined effects of environmental, pathological, and physiological factors 6,11,12 . Thus, early treatment should not just be symptomatic use of antipsychotic drugs such as haloperidol, olanzapine, and aripiprazole, but should also involve etiological treatments to reduce pain and suffering, correct hypoproteinemia and anemia, restore water and electrolyte balance, provide adequate oxygen inhalation and nutritional support, ensure good sleep, and allow patients to carry out early functional exercises.
In addition, early surgery may be an important measurement to prevent delirium and reduce mortality. The results of this study showed that the delirium group had a longer injury-to-surgery time than the nondelirium group. Pioli et al. 18 considered the delay of surgery to be a risk factor for delirium. For older adult hip fracture patients with mild to moderate cognitive impairment, early surgery should be one of the preoperative goals. Other scholars have reported that early surgical treatment, epidural anesthesia, and a shortened hospital stay can reduce mortality following hospitalization of elderly patients with hip fractures 3,19,20 . Therefore, in view of the characteristics of hip fractures in the older adult, active development of geriatric wards, multidisciplinary cooperation, and shortening the waiting time for surgery will help reduce POD and improve the early prognosis of patients.
This study has three limitations. First, although the gold standard for diagnosis of delirium, the CAM assessment, was used in this study and determined by individuals quali ed as associate chief physician or above, the core symptoms of delirium are di cult to de ne and clinical symptoms, severity, and progression, are di cult to identify. As such, the CAM diagnosis method is based on the subjective judgment of the clinician and not quantitative scores 10 , so there is a certain degree of subjective bias. Second, a valid malnutrition diagnostic measure (eg MNA / MNA-SF) was not conducted in this study. In addition, as the duration, classi cation, and severity of delirium were not observed in this study, a prospective multicenter cohort study will be used for further analysis.

Conclusion
POD has a higher incidence in older adult hip fracture patients. Compared with the non-delirium group, the delirium group had a longer waiting time before surgery, was more prone to perioperative complications, had a longer hospital stay, and had a higher mortality rate within 1 year. Thus, early detection, early diagnosis, and early treatment of delirium after surgery have important clinical signi cance.