As shown in Fig. 1, 874 elderly patients with hip fractures were initially included in this study, of which 358 patients were excluded due to: old fractures (43), incomplete data or abnormal data (65), multiple injuries (34), pathological fractures (2), and isolated calf intermuscular thrombosis (214). Finally, a total of 516 patients were included.
Forty-four patients developed DVT within 7 days after hip fracture and 20 patients had DVT after 1 week. Of the 64 patients, 41 patients developed DVT in the injured limb, 11 patients (17.2%) had DVT in the uninjured limb, and 12 (18.8%) cases of DVT presented in bilateral lower limbs. In addition, forty-seven (73.4%) patients had two or more sites of DVT. (Table 1).
Table 1 The predilection site and time of deep venous thrombosis
Occurrence time of DVT
|
1-3 days
|
4-7 days
|
>7 days
|
Total (n)
|
Locations of DVT
|
|
|
|
|
Posterior tibial vein
|
8(34.78)
|
8(38.10)
|
8(36.36)
|
24
|
Peroneal vein
|
12(52.17)
|
13(61.90)
|
12(54.55)
|
37
|
Popliteal vein
|
3(13.04)
|
5(23.81)
|
8(36.36)
|
16
|
External iliac vein
|
0(0.0)
|
0(0.0)
|
1(4.54)
|
1
|
Common femoral vein
|
2(8.69)
|
2(9.52)
|
2(9.09)
|
6
|
Superficial femoral vein
|
5(21.74)
|
5(23.81)
|
6(27.27)
|
16
|
Profunda femoral vein
|
0(0.0)
|
2(9.52)
|
4(18.18)
|
6
|
Locations of DVT
|
|
|
|
|
Injured lower limb
|
18(78.26)
|
10 (52.38)
|
13(54.55)
|
41 (64.1)
|
Contralateral lower limb
|
4 (17.39)
|
5 (28.57)
|
2(13.64)
|
11 (17.2)
|
Bilateral lower limb
|
1 (4.35)
|
6(19.04)
|
5(22.73)
|
12 (18.8)
|
Total (n)
|
23(35.94)
|
21(32.81)
|
20(31.25)
|
64
|
Abbreviation: DVT, deep vein thrombosis.
Notes: Values are presented as the number (percentage).
The predilection sites of DVT were the peroneal vein (52.17%, 12/23), posterior tibial vein (38.10%, 8/23), and superficial femoral vein (21.74%, 5/23) within 3 days after fracture. Patients were more prone to develop DVT in the peroneal vein (54.55%,12/20), popliteal vein, (36.36%,8/20), and posterior tibial vein (36.36%,8/20), after 1 week of fracture. The peroneal vein was the most vulnerable site for DVT at different periods. (Table 1).
Univariate analysis showed that age (81.6 ± 7.0 vs 79.1 ± 7.9, P = 0.017), ACCI (5.5 ± 1.3 vs 4.8 ± 1.4, P = 0.001), Intertrochanteric fracture (67.2% vs 48.5%, P = 0.005), hypoproteinemia(48.4% vs 31.2%, P = 0.006), hypertension(67.2% vs 50.0%, P = 0.010), HB at admission (108.3 ± 15.9 vs 113.9 ± 17.5, P = 0.015), the time interval between fracture and admission (4.2 ± 4.6 vs 2.1 ± 3.5, P = 0.001), the time interval between fracture and operation (9.6 ± 5.8 vs 6.9 ± 4.4, P = 0.001), and D-dimer at admission (4.5 ± 3.9 vs 3.8 ± 4.3, P = 0.007) were significant risk factors for postoperative DVT (Table 2).
Table 2
The comparison between the two groups
Variables | DVT group | Non-DVT group | P |
Age (years) | 81.6 ± 7.0 | 79.1 ± 7.9 | 0.017 * |
BMI (Kg/m2) | 24.3 ± 3.8 | 23.6 ± 3.9 | 0.148 |
ACCI | 5.5 ± 1.3 | 4.8 ± 1.4 | 0.001 * |
Gender, n (%) | | | 0.289 |
Male | 15 (23.4) | 135 (70.1) | |
Female | 49 (76.6) | 317 (29.9) | |
Fracture site, n (%) | | | 0.005 * |
Intertrochanteric fracture | 43 (67.2) | 219 (48.5) | |
Femoral neck fracture | 21 (32.8) | 233 (51.5) | |
Injury mechanism, n (%) | | | 0.739 |
Low energy damage | 62 (96.9) | 434 (96) | |
High energy damage | 2 (3.1) | 18 (4) | |
Infection, n (%) | 14 (21.9) | 68 (15) | 0.162 |
Cognitive impairment, n (%) | 0 (0) | 14 (3.1) | 0.153 |
Anemia, n (%) | 31 (48.4) | 197 (43.8) | 0.483 |
Hypertension, n (%) | 43 (67.2) | 226 (50.0) | 0.010 * |
Diabetes, n (%) | 19 (29.7) | 134 (29.6) | 0.995 |
Coronary disease, n (%) | 20 (31.3) | 99 (21.9) | 0.097 |
Cerebrovascular disease, n (%) | 37 (57.8) | 224 (49.6) | 0.216 |
Respiratory disease, n (%) | 10 (15.6) | 95 (21.0) | 0.316 |
Cancer, n (%) | 4 (6.3) | 16 (3.5) | 0.295 |
Hypoalbuminemia, n (%) | 31(48.4) | 141(31.2) | 0.006 * |
D-dimer at admission ≥ 2.5 mg/L | 44 (68.8) | 217(48.0) | 0.002 * |
The time from injury to admission (days) | 4.2 ± 4.6 | 2.1 ± 3.5 | 0.001 * |
The time from injury to operation (days) | 9.6 ± 5.8 | 6.9 ± 4.4 | 0.001 * |
HB at admission (g/L) | 108.3 ± 15.9 | 113.9 ± 17.5 | 0.015 * |
Cholesterol at admission | 3.9 ± 0.9 | 4.0 ± 0.9 | 0.140 |
D-dimer at admission (mg/L) | 4.5 ± 3.9 | 3.8 ± 4.3 | 0.007* |
Abbreviation: BMI, body mass index; ACCI: age-adjusted Charlson Comorbidity Index; HB, hemoglobin. |
Note: Values are mean ± standard deviation or number (percentage) as appropriate. *P < 0.05, statistical signifificance. In this study, anemia was defined as hemoglobin level < 120g/L for males and < 110g/L for females.
The ROC analysis showed that the AUC of age, ACCI, HB at admission, the time interval between fracture and admission, the time interval between fracture and operation, and D-dimer at admission were 0.60 (95% CI, 0.53 to 0.68; P = 0.008), 0.65 (95% CI, 0.58 to 0.72; P = 0.001), 0.40 (95% CI, 0.33 to 0.47; P = 0.011), 0.67 (95% CI, 0.60 to 0.75; P = 0.001), 0.65 (95% CI, 0.58 to 0.73; P = 0.001), and 0.60 (95% CI, 0.54 to 0.67; P = 0.008), respectively. The optimal cut-off values were 78 years old, 6.0, 75.3 g/L, 1.0 days, 6.0 days, and 2.5 mg/L respectively (Table 3).
Table 3
ROC curve analysis results of statistically significant continuous variables
Variables | AUC (95%CI) | Cutoff value | Youden index | Standard error | Sensitivity | Specificity | P-value |
Age (years) | 0.60 (0.53, 0.68) | 77.50 | 0.198 | 0.036 | 76.6 | 43.2 | 0.008 |
ACCI | 0.65 (0.58, 0.72) | 5.50 | 0.250 | 0.036 | 51.6 | 73.4 | 0.001 |
HB at admission(g/L) | 0.40 (0.33, 0.47) | 75.3 | 0.006 | 0.011 | 98.4 | 2.2 | 0.011 |
The time from injury to admission | 0.67 (0.60, 0.75) | 1.3 | 0.317 | 0.038 | 59.4 | 72.3 | 0.001 |
The time from injury to operation | 0.65 (0.58, 0.73) | 6.5 | 0.252 | 0.039 | 62.50 | 62.7 | 0.001 |
D-dimer at admission(mg/L) | 0.60 (0.54, 0.67) | 2.5 | 0.207 | 0.035 | 68.8 | 51.9 | 0.008 |
Abbreviation: BMI, body mass index; ACCI, age-adjusted Charlson Comorbidity Index; CRP, C-reaction protein; ROC, receiver operating characteristic; AUC, the area under the curve; 95%CI, 95% confidence interval. |
The result of multivariate analysis showed that hypertension (OR, 2.5, 95%CI, 1.4 to 4.7; P = 0.003), ACCI ≥ 6.0 (OR, 1.4, 95%CI, 1.1to 1.8; P = 0.009), D-dimer ≥ 2.5mg/L at admission (OR, 3.1, 95%CI, 1.7 to 5.9; P = 0.001), the time interval between fracture and admission ≥ 1 day (OR, 1.1, 95%CI, 1.0 to 1.2; P = 0.003), and the time interval between fracture and operation ≥ 6 days (OR, 1.1, 95%CI, 1.0 to 1.2; P = 0.002) were independent risk factors for preoperative DVT in elderly patients with hip fracture (P < 0.05) (Table 4).
Table 4
Multivariate logistic regression analysis of deep vein thrombosis
Variables | Regression coefficient | Standard error | Wald χ2 value | OR (95%CI) | P |
Hypention | 0.927 | 0.317 | 8.538 | 2.5 (1.4,4.7) | 0.003 |
The time from injury to admission | 0.108 | 0.036 | 9.046 | 1.1 (1.0,1.2) | 0.003 |
The time from injury to operation | 0.085 | 0.027 | 9.643 | 1.1 (1.0,1.2) | 0.002 |
ACCI | 0.324 | 0.124 | 6.788 | 1.4 (1.1,1.8) | 0.009 |
D-dimer at admission ≥ 2.5mg/L | 1.137 | 0.322 | 12.464 | 3.1 (1.7,5.9) | 0.001 |
Abbreviation: ACCI, age-adjusted Charlson Comorbidity Index; OR, odd ratio; 95%CI, 95% confidence interval.