Risk factors for preoperative deep venous thrombosis in hip fracture patients: a meta-analysis

Study identification and selection

Initially, we collected a total of 160 English and 51 Chinese articles through the database search. Of these, 105 English articles and 26 Chinese articles were excluded due to repetition, and 22 English articles and 6 Chinese articles were removed from review on the basis of the titles and abstracts. The remaining 33 English articles and 19 Chinese articles were retrieved for inclusion criteria, and 19 English articles and 7 Chinese articles were excluded. Finally, 14 English articles and 12 Chinese articles that met our inclusion criteria were included in the present meta-analysis. The selection process that was used in this meta-analysis is shown in Fig. 1.

Fig. 1figure 1

Flow diagram of study selection

Baseline characteristics and quality assessment

The main characteristics of the 26 English articles (9823 patients) published before January 2022 and included in the meta-analysis are presented in Table 1.

Table 1 Characteristics of included studies

Because all studies included were retrospective studies, we used the Newcastle Ottawa Quality Assessment Scale (NOQAS) to assess the quality of each study. This scale for non-randomized case-controlled studies and cohort studies was used to allocate a maximum of nine points for the quality of selection, comparability, exposure, and outcomes for study participants. Twenty of these studies scored eight points, and another six scored seven points. Hence, the quality of each study was relatively high (Table 2).

Table 2 The quality assessment according to the Newcastle Ottawa Quality Assessment Scale (NOQAS) of each studyAge

Sixteen studies [10, 11, 16,17,18,19,20,21,22,23,24,25,26,27,28,29] reported the relationship between age at surgical time and preoperative DVT. The test for heterogeneity was not significant and the studies had low heterogeneity (p for heterogeneity = 0.52, I2 = 0%, Fig. 2a and Table 3). In this study, advanced age at surgical time was a risk factor for preoperative DVT [fixed-effects model; p = 0.0003, OR = 0.13, 95% CI (0.06, 0.21), Fig. 2a and Table 3]. Furthermore, we analyzed subgroups of age. The results indicated that patients over the age of 90 had an increased risk of preoperative DVT compared with other age groups, as shown in Fig. 2b–d and Table 3. From Fig. 2e–g and Table 3, no significant difference was found among other groups.

Fig. 2figure 2

Forest plot showing age in 2 groups. CI confidence interval, df degrees of freedom, M-H Mantel–Haenszel. a Relationship between age at surgical time and preoperative DVT; b 60–70 years vs > 90 years; c 70–80 years vs > 90 years; d 80–90 years vs > 90 years; e 60–70 years vs 70–80 years; f 60–70 years vs 80–90 years; g 70–80 years vs 80–90 years

Sex

Twenty-six studies [10, 11, 14, 16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38] reported the relationship between sex and preoperative DVT. The test for heterogeneity was not significant and the studies had low heterogeneity (p for heterogeneity = 0.16, I2 = 22%). In this study, female patient was a risk factor for preoperative DVT [fixed-effects model; p = 0.0009, OR = 0.82, 95% CI (0.72, 0.92)] (Fig. 3 and Table 3).

Fig. 3figure 3

Forest plot showing sex in 2 groups. CI confidence interval, df degrees of freedom, M-H Mantel–Haenszel

Body mass index

Twelve studies [10,11,12, 19,20,21, 25, 27, 30,31,32,33] reported the relationship between BMI at surgical time and preoperative DVT. The test for heterogeneity was not significant and the studies had low heterogeneity (p for heterogeneity = 0.92, I2 = 0%, Fig. 4a and Table 3). In this study, BMI at surgical time was not a risk factor for preoperative DVT [fixed-effects model; p = 0.19, OR = 0.07, 95% CI (−0.03, 0.17), Fig. 4a and Table 3]. Furthermore, we analyzed subgroups of BMI. The results indicated that patients with more than 28 kg/m2 had an increased risk of preoperative DVT compared with the less than 18.5 kg/m2 group and 24.0–27.9 kg/m2 group, shown in Fig. 4b, c and Table 3. From Fig. 4d–h and Table 3, there was no significant difference among other groups.

Fig. 4figure 4

Forest plot showing body mass index in 2 groups. CI confidence interval, df degrees of freedom, M-H Mantel–Haenszel. a Relationship between BMI at surgical time and preoperative DVT; b  < 18.5 kg/m2 vs > 28 kg/m2; c 24.0–27.9 kg/m2 vs > 28 kg/m2; d  < 25 kg/m2 vs > 25 kg/m2; e  < 18.5 kg/m2 vs 18.5–23.9 kg/m2; f  < 18.5 kg/m2 vs 24.0–27.9; g 18.5–23.9 kg/m2 vs 24.0–27.9 kg/m2; h 18.5–23.9 kg/m2 vs > 28 kg/m2

Type of hip fracture

Fourteen studies [10, 11, 16, 17, 23, 24, 26, 27, 29, 30, 34,35,36, 38] reported the relationship between type of hip fracture and preoperative DVT. Intertrochanteric fracture vs femur neck fracture: p for heterogeneity = 0.89, I2 = 0%; Intertrochanteric fracture vs subtrochanteric fracture: p for heterogeneity = 0.53, I2 = 0%; Femur neck fracture vs subtrochanteric fracture: p for heterogeneity = 0.48, I2 = 0%). In this study, patients with subtrochanteric fracture had the highest rate of preoperative DVT, while femur neck fracture had the lowest rate [intertrochanteric fracture vs femur neck fracture, fixed-effects model; p < 0.0001, OR = 1.43, 95% CI (1.20, 1.72); intertrochanteric fracture vs subtrochanteric fracture, fixed-effects model; p = 0.0003, OR = 0.28, 95% CI (0.14, 0.55); femur neck fracture vs subtrochanteric fracture, fixed-effects model; p = 0.0002, OR = 0.34, 95% CI (0.19, 0.60)] (Fig. 5 and Table 3).

Fig. 5figure 5

Forest plot showing type of fracture in 2 groups. CI confidence interval, df degrees of freedom, M-H Mantel–Haenszel. a Intertrochanteric fracture vs femur neck fracture; b intertrochanteric fracture vs subtrochanteric fracture; c femur neck fracture vs subtrochanteric fracture

Time from injury to surgery

Six studies [11, 19, 21, 34,35,36] reported the relationship between time from injury to surgery and preoperative DVT. The test for heterogeneity was not significant and the studies had low heterogeneity (p for heterogeneity = 0.76, I2 = 0%). In this study, prolonged time from injury to admission was a risk factor for preoperative DVT [fixed-effects model; p < 0.00001, OR = 2.06, 95% CI (1.40, 2.72)]. Moreover, we also explored whether 5 days as a cut-off time affected the rate of preoperative DVT. The test for heterogeneity was not significant and the studies had low heterogeneity (p for heterogeneity = 0.68, I2 = 0%). In this study, ≥ 5 days from injury to admission had a significantly higher rate of preoperative DVT compared with < 5 days [fixed-effects model; p < 0.00001, OR = 4.54, 95% CI (2.50, 8.25)] (Fig. 6 and Table 3).

Fig. 6figure 6

Forest plot showing prolonged time from injury to admission in 2 groups. CI confidence interval, df degrees of freedom, M-H Mantel–Haenszel. a Prolonged time from injury to admission; b  ≥ 5 days vs < 5 days

Time from injury to admission

Five studies [18, 19, 21, 31, 33] reported the relationship between time from injury to admission and preoperative DVT. The test for heterogeneity was not significant and the studies had low heterogeneity (p for heterogeneity = 0.49, I2 = 0%). In this study, prolonged time from injury to admission was a risk factor for preoperative DVT [fixed-effects model; p < 0.00001, OR = 0.54, 95% CI (0.44, 0.65)] (Fig. 7 and Table 3).

Fig. 7figure 7

Forest plot showing time from injury to admission in 2 groups. CI confidence interval, df degrees of freedom, M-H Mantel–Haenszel

Location

Two studies [21, 32] reported the relationship between the location of living place (rural or city) and preoperative DVT. The test for heterogeneity was not significant and the studies had low heterogeneity (p for heterogeneity = 0.36, I2 = 0%). In this study, city location was not a risk factor for preoperative DVT [fixed-effects model; p = 0.75, OR = 1.07, 95% CI (0.72, 1.58)] (Fig. 8 and Table 3).

Fig. 8

留言 (0)

沒有登入
gif