*2.1. Study Population*

This retrospective comparative study was performed in a single institution and approved by our institutional review board. The data were collected from medical records between March 2018 and April 2021. The inclusion criteria were as follows: (1) age of ≥60 years and (2) surgical treatment of proximal femoral fracture, which was defined as a femoral neck fracture, an intertrochanteric femoral fracture (AO/OTA 31) [10], and a subtrochanteric fracture (fracture extending 5 cm below the lower border of the lesser trochanter) [11]. Meanwhile, the exclusion criteria were as follows: (1) previous psychiatric illness, (2) previous cognitive impairment, (3) pathologic fracture, (4) prophylactic fixation, (5) revision of total hip replacement, (6) delayed surgery with neglected fracture, and (7) incomplete clinical data. Initially, 307 patients were included; ultimately, 252 patients were enrolled in this study. The patients who received general or spinal anesthesia followed by RNB were grouped into the RNB group. The patients who received only general or spinal anesthesia were grouped into the control group (Figure 1).

**Figure 1.** Flow diagram of patient enrollment and grouping.

#### *2.2. Procedures*

General or spinal anesthesia was performed according to the patients' overall health status, and RNB use was left to the discretion of the anesthesiologist. Under ultrasound guidance, a single shot of RNB, including fascia iliaca compartment block (FICB) [12], lumbar plexus block (LPB) [13], and pericapsular nerve group (PENG) block [14], was applied (Figure 2). For ultrasound-guided RNBs, a transportable ultrasound with 60 mm convex 2–5 MHz transducer for LPB and 25 mm linear 18–4 MHz transducer (Sonimage HS1TM, Konica Minolta Inc., Wayne, NJ, USA), and 21-gauge echoplex needle (Vygon, Ecouen, France) were used. After determining the insertion site, real-time ultrasoundguided perineural injection was conducted with the patient in supine or lateral position. Then, 30 mL of 0.3% ropivacaine was administered for LPB and FICB. Additionally, 20 mL of 0.3% ropivacaine was administered for PENG block.

**Figure 2.** Illustration of the target nerves according to the type of regional nerve block. The fascia iliaca compartment block (FICB) targets the lateral femoral cutaneous (LFCN), femoral (FN), and obturator nerves (ON). The lumbar plexus block (LPB) targets the lumbar plexus (FN, ON, and LFCN) (dotted circle). The pericapsular nerve group (PENG) block mainly targets the articular branches (black arrowhead) of the FN and accessory obturator nerve (AON) (blue solid circle). SN: Sciatic nerve.

The patients with femoral neck fracture underwent bipolar hemiarthroplasty or internal fixation with multiple screws, while those with intertrochanteric and subtrochanteric fractures underwent intramedullary nailing. After surgery, intravenous patient-controlled analgesia (PCA) was applied in all patients. On the day of surgery, intravenous acetaminophen 1 g once a day was administered unless the patient had a contraindication. If the weight of the patient was less than 50 kg, the dose of acetaminophen was controlled at 15 mg/kg. The day after surgery, oral pain medications included a tramadol 37.5 mg/acetaminophen 325 mg tablet given twice daily and tapentadol 100 mg given twice daily. Rescue analgesia during postoperative period was intravenous tramadol 50 mg or hydromorphone 0.5 or 1.0 mg. Periarticular injection was not performed. We encouraged wheelchair ambulation as soon as possible from the day after surgery, and tolerable weight-bearing ambulation started on 2 days postoperatively.

#### *2.3. Data Collection*

Demographic data were collected from the patients' medical records, including age, sex, body mass index (BMI), age-adjusted Charlson comorbidity index (ACCI), Koval score before injury, injury mechanism, anesthesia method, fracture type, time from injury to

surgery, time from admission to surgery, and type of surgery. The Charlson comorbidity index (CCI) was calculated by adding the coefficient assigned to comorbidities when injured. The ACCI was calculated by adding 1 point for each decade after the age of 40 years to the CCI value [15]. Falls from heights of 1 m or less were defined as "low-energy mechanism of injury" [16].

Delirium was recorded and defined using the confusion assessment method on any postoperative day or night of their hospital stay following surgery as the primary outcome [17]. Postoperative pain score, pain medication consumption, and rehabilitation assessment were evaluated as the secondary outcomes. The pain scores at 6, 12, 24, and 48 h postoperatively were assessed using the visual analog scale pain score. The consumption of pain medication for 2 postoperative days was examined. The amount of analgesics was calculated into milligrams of oral morphine according to the equianalgesic table [18–20]. A rehabilitation assessment was performed on the basis of the time (days) from surgery to wheelchair ambulation. All these parameters (pain score, incidence of delirium, pain medication consumption, and rehabilitation assessment) were included in the standardized protocol of our hospital.

The primary and secondary outcomes were compared between the RNB and control groups. A subgroup analysis for incidence of postoperative delirium between general and spinal anesthesia was performed. A subgroup analysis between FICB, LPB, and PENG block was performed to evaluate the differences between the blocks.

A multivariable analysis was performed to assess which variables were associated with incidence of delirium and clinically significant parameters were included in the model such as age, gender, CCI, RNB, and anesthesia type [21–23]. Nerve block-related complications, such as falls within 48 h after surgery and nerve injury, were assessed from the medical records. Postoperative medical complications, including deep vein thrombosis, pulmonary embolism, pneumonia, angina, myocardial infarction, and urinary tract infection, were evaluated.

#### *2.4. Statistical Analysis*

Categorical variables (e.g., sex, injury mechanism, method of anesthesia, fracture type, type of surgery, and incidence of delirium) were analyzed using the chi-square or Fisher's exact test. Continuous variables (e.g., age, BMI, ACCI, Koval score, time from injury to surgery, pain score, amount of analgesics consumed, and time to wheelchair ambulation) were analyzed using an independent t-test or the Mann–Whitney test. The Shapiro–Wilk test was used to check if the data distribution was normal. The Kruskal–Wallis test was used when compare more than two continuous variables. A logistic regression analysis was conducted for multivariable analysis. All continuous data are described as means and standard deviations. Statistical significance was accepted for *p*-values of <0.05 using SPSS version 23.0 (IBM Corp., Armonk, NY, USA).

#### **3. Results**

The RNB group consisted of 129 patients, while the control group consisted of 123 patients. The baseline characteristics of the two study groups are presented in Table 1. There was no significant difference found in the average age, BMI, ACCI, Koval score before injury, injury mechanism, method of anesthesia, fracture type, time from injury to surgery, time from admission to surgery, and type of surgery. The methods of RNB are summarized in Table 2.


**Table 1.** Baseline characteristics of the RNB and control groups.

RNB, regional nerve block; BMI, body mass index; ACCI, age-adjusted Charlson comorbidity index; \* calculated from 249 patients who had clinical records on the preoperative Koval score; † calculated from 246 patients who had clinical records on the time from injury to surgery.

**Table 2.** Type of RNB for hip fracture.


RNB, regional nerve block.

The overall incidence of delirium was 21%; the rate in the RNB group was lower than that in the control group (15 vs. 27%, *p* = 0.027). There were no nerve block-related complications, including falls within 48 h after surgery and nerve injury.

The average pain score at 6 h postoperatively in the RNB group was lower than that in the control group (2.8 ± 1.5 vs. 3.3 ± 1.6, respectively, *p* = 0.030) (Table 3). There was no significant difference in the pain score at 12 (RNB group: 2.8 ± 1.6 vs. control group: 2.7 ± 1.5, *p* = 0.432), 24 (RNB group: 2.5 ± 1.3 vs. control group: 2.4 ± 1.4, *p* = 0.154), and 48 (RNB group: 2.2 ± 1.1 vs. control group: 2.0 ± 1.1, *p* = 0.083) hours postoperatively. There was also no significant difference in the amount of opioids consumed for 2 postoperative days (oral morphine; RNB group: 36.2 ± 25.4 mg vs. control group: 33.5 ± 28.0 mg, *p* = 0.322) and time from injury to wheelchair ambulation (RNB group: 1.8 ± 2.4 days vs. control group: 1.7 ± 1.1 days, *p* = 0.407).

In the subgroup analysis between general and spinal anesthesia, there was no significant difference in the incidence of postoperative delirium (Table 4). In the subgroup analysis between FICB, LPB, and PENG block, there was no significant difference in the incidence of postoperative delirium, pain score, and amount of opioids consumed, but time to wheelchair ambulation presented a significant difference for PENG block compared with FICB and LPB (Table 5).


**Table 3.** Comparison of the postoperative pain score, incidence of delirium, amount of opioids consumed, and time to wheelchair ambulation.

RNB, regional nerve block; \* expressed as milligrams of oral morphine by equianalgesic conversion; † calculated from 247 patients who had clinical records on the time to wheelchair ambulation.

**Table 4.** Subgroup analysis for the incidence of delirium between general and spinal anesthesia.


RNB, regional nerve block.

**Table 5.** Subgroup analysis for the postoperative pain score, incidence of delirium, amount of opioids consumed, and time to wheelchair ambulation between FICB, LPB and PENG block.


FICB, fascia iliaca compartment block; LPB, lumbar plexus block; PENG block, pericapsular nerve group block; \* expressed as milligrams of oral morphine by equianalgesic conversion; † calculated from 106 patients who had clinical records on the time to wheelchair ambulation.

In the multivariable analysis, age and RNB were significantly associated with the incidence of postoperative delirium (Table 6). There was also no significant difference in the postoperative medical complications between them (Table 7).

**Table 6.** Bivariate and multivariable logistic regression analysis for the risk factors of postoperative delirium.


\* The results of multivariable logistic regression analysis were presented only for the variables which were remained in the final model.


**Table 7.** Postoperative complications.

RNB, regional nerve block.

#### **4. Discussion**
