DOI: https://doi.org/10.21203/rs.3.rs-1778395/v2
Objective: This study aimed to find out the risk factors of postoperative moderate anemia (PMA) to develop a scoring scale for predicting the occurrence of PMA and to determine the recommended preoperative hemoglobin level in spinal tuberculosis (STB) patients.
Methods: A total of 223 STB patients who underwent focus debridement from January 2012 to March 2020 were enrolled in the study. The study cohort was divided into two groups owing to the occurrence of PMA. Moderate anemia was defined as a hemoglobin level of < 90 g/L. The clinical characteristics of STB patients who developed PMA were evaluated, and a scale was developed by logistic regression analysis. The performance of this scoring scale is prevalidated.
Results: Of the 223 patients, 76 developed PMA. Multivariate binary logistic regression analysis showed that body mass index, diabetes, low preoperative hemoglobin level, long operation time, and posterior approach were independent risk factors for PMA in STB patients. These significant items were assigned scores to create a scoring scale as to predicting PMA, and receiver operating characteristic (ROC) curve analysis implicated that the optimal cutoff score was 4 points. On the basis of the scoring scale, patients with scores within 0–3 points were defined as the low-risk group; those with scores within 4–6 points were defined as the moderate-risk group; and those with scores within 7–10 points were defined as the high-risk group. The perioperative decrease in hemoglobin level was 20.07 ± 10.47 g/L in the low-risk group, 24.44 ± 12.67 g/L in the moderate-risk group, and 29.18 ± 10.34 g/L in the high-risk group.
Conclusion: According to the scoring scale, patients with STB with a score of 0–3 points have a low risk of PMA, those with a score of 4–6 have a moderate risk, and those with a score of 7–10 have a high risk. The recommended preoperative hemoglobin levels for the low-, moderate-, and high-risk groups are 110, 115, and 120 g/L, respectively.
Spinal tuberculosis (STB) is a common extrapulmonary tuberculosis that accounts about approximately half in all osteoarticular tuberculosis cases (1). Currently, it is considered that antituberculous drugs combined with surgery is the gold standard for the treatment of STB (2). Lesion focus debridement is a significant treatment in STB therapy, which enhances tuberculosis control, promotes bone graft fusion, improves the efficacy of antituberculosis drugs, and reduces the risk of STB recurrence (3, 4).
However, debridement of the spinal focus is an iatrogenic trauma that usually causes considerable blood loss. It has been reported that 42% of patients aged > 60 years who underwent spinal surgery were malnourished preoperatively (5, 6). Patients with STB are susceptible to the development of postoperative anemia because of intraoperative blood loss and increased inflammatory cytokines after debridement, which induce a response characterized by effects such as the uptake of iron in the gastrointestinal tract decreased, the iron sequestration in macrophages decreased, the erythroid response to erythropoietin diminished, and the production of erythropoietin decreased (7–9). Postoperative moderate anemia (PMA) in patients with STB can lead to various adverse outcomes. Studies have reported that in the presence of anemia, not only the leukocyte bactericidal ability but also the cell-mediated immune response are obviously inhibited, which has adverse effects on postoperative infection control and tuberculosis treatment in patients with STB (10, 11). In addition, postoperative anemia is closely related to surgical site infections, unsatisfying physical function, poor recovery, prolonged hospital stay, and mortality, and PMA has a higher risk of causing adverse events than mild and severe anemia (8, 12, 13). Because of the characteristics and potential hazards of PMA in patients with STB, reducing the occurrence of PMA in this patient population has great clinical significance. Therefore our study is to identify the risk factors of PMA in STB patients, and on this basis, to establish the recommended preoperative hemoglobin values for patients with different risk groups of PMA.
All study participants provided written informed consent to store their data in a hospital database and use it for research purposes. This work was reported according to the STROCSS (Strengthening the Reporting of Cohort Studies in Surgery) criteria.
This retrospective study included 223 STB patients who underwent focus debridement from January 2012 to March 2020.
Patients were selected when they met these inclusion criteria: (i) the medical records is totally complete, including general information, perioperative laboratory data, imaging data (magnetic resonance imaging and computed tomography), and data on perioperative clinical features; (ii) underwent the lesion debridement surgery; and (iii) postoperative pathological diagnosis of STB.
Patients were excluded when they met these exclusion criteria: (i) STB was suspected but not confirmed with pathological examination, (ii) preliminary and pathological diagnosis of a disease other than STB, (iii) accepted the conservative treatment, and (iv) a history of previous STB.
On the basis of our clinical experience and previous researches, the following possible predictors of PMA in STB patients were included: general patients condition, laboratory examination indices, imaging examination indices, and surgery-related indices. Measures of general patient condition included age, sex, body mass index (BMI), comorbidities, history of alcohol drinking, history of cigarette smoking, and disease course. Laboratory examination indices included preoperative hemoglobin level, preoperative serum albumin level, preoperative lymphocyte count, preoperative erythrocyte sedimentation rate, preoperative C-reactive protein level, and daily hemoglobin level during 5 days after surgery. Imaging examination indices included the number of diseased vertebrae, number of fixation segments, and number of pedicle screws. Surgery-related indices included operation time, surgical approach, and operation blood loss. The postoperative moderate anemia was defined as a postoperative hemoglobin level of < 90 g/L.
All enrolled patients were divided into two groups according to the occurrence of PMA. To identify possible predictors of PMA, the general condition, laboratory examination index, imaging examination index and operation related index of the two groups were analyzed by univariate analysis. Then multivariate logistic regression analysis was performed to determine the predictors of PMA in patients with STB. These items are assigned to establish a scoring scale.
The continuous variables with statistical significance (BMI, preoperative hemoglobin, operation time) were converted into binary variables through the analysis of receiver operating characteristic (ROC) curve. And based on the the method used by Kharbanda et al. and our previous study, the factors with P < 0.05 in the multivariate logistic regression analysis were included in the final scale (14, 15). The weighted score of each item was determined according to the relative size of the P-value.
Find out the appropriate cut-off point of the scoring scale by determining the point closest to the upper left corner of the ROC curve.
Between April 2020 and March 2021, a total of 72 STB patients were included to validate the accuracy of the scoring scale. The inclusion and exclusion criteria were identical in both validation set and derivation set.
All enrolled patients were followed up by clinicians using phone calls and outpatient consultations. All medical data were complete in included patients.
ROC curve analysis determined the optimal cutoff point of the scoring scale and the threshold values of continuous variables. The prevalence rate of clinical features was evaluated by calculating the sensitivity and specificity of various factors. The clinical features were first entered into univariate logistic regression analyses to find out the significant factors, and then the significant factors were entered into multivariate logistic regression analyses. The significant factors of the scoring scale were determined by multivariate logistic regression, and the weighted score of each factor was on account of the relative size of the P-value. Statistical significance was defined as P < 0.05. SPSS software (version 26.0; IBM Corporation, Armonk, New York, USA) was used for all statistical analyses.
3.1 Patient population
Among the 223 patients included in the study, 76 presented with PMA (27 men and 49 women). The remaining 147 patients (99 men and 48 women) did not develop PMA (Table 1). The average age of patients with PMA was 52.26 ± 16.81 and that of patients without PMA was 44.84 ± 15.46 years (Table 1).
3.2 Results of logistic regression analyses
Univariate logistic regression analyses revealed that age, sex, BMI, diabetes mellitus, disease course, operation blood loss, operation time, surgical approach, number of fixation segments, preoperative hemoglobin level, preoperative lymphocyte count, preoperative serum albumin level, and preoperative erythrocyte sedimentation rate were risk factors for PMA (Table 2). Multivariate logistic regression analyses of the identified significant factors revealed that BMI, diabetes mellitus, preoperative hemoglobin, operation time, and surgical approach were independent risk factors for PMA (Table 3). ROC curves showed that the optimal cutoff value of BMI was 19.5 kg/m2 (sensitivity, 0.721; specificity, 0.513), the cutoff value of preoperative hemoglobin was 120 g/L (sensitivity, 0.755; specificity, 0.737), and the cutoff value of operation time was 205 min (sensitivity, 0.605; specificity, 0.612) (Figure 1).
3.3 Development of the scoring scale
The five clinical characteristics identified in multivariate analyses were assigned to establish the scoring scale. According to the P-values, BMI (P = 0.029) and posterior surgical approach (P = 0.046) were assigned 1 point each, operation time (P = 0.001) was assigned 2 points, and diabetes mellitus (P < 0.001) and preoperative hemoglobin level (P < 0.001) were assigned 3 points each (Table 4). The ROC curve showed that 4 points is the optimal cutoff score of the scoring scale (sensitivity, 0.855; specificity, 0.701) (Figure 2).
3.4 Validation of the scoring scale
The scoring scale was prospectively used to the 72 patients in the validation set. The comparison of the results of the scoring scale applied to both derivation and validation set is shown in Table 5. On the basis of the cutoff score of 4 points, the sensitivity and specificity of the scoring scale for predicting PMA in STB patients were 59.6% and 90.4% in the derivation set and 58.5% and 90.3% in the validation set, respectively (Table 5).
3.5 Risk prediction using the scoring scale
Using the collected medical data of the 223 patients and the scoring scale, patients were scored and divided into three groups: 125 patients with scores within 0–3 points were defined as the low-risk group; 121 patients with scores within 4–6 points were defined as the moderate risk group; and 53 patients with scores within 7–10 points were defined as the high-risk group. The incidence of PMA was 9.7%, 49.4%, and 89.3% in the low-, moderate-, and high-risk groups, respectively (Table 6).
3.6 Recommended preoperative hemoglobin level for the three groups
The trends of hemoglobin change during 5 days after surgery were observed in all patients (Figure 3) and in each of the low-, moderate-, and high-risk groups (Figure 4). By combining the diagnostic threshold of moderate anemia and the respective average decrease in hemoglobin level in the three groups (Table 7), we determined the recommended preoperative hemoglobin level for patients with STB (Table 8).
In our study, multivariate logistic regression analysis revealed that BMI, diabetes mellitus, preoperative hemoglobin level, operation time, and posterior approach were independent risk factors for PMA in STB patients.
Diabetes was found as an independent risk factor for PMA in STB patients. We observed that STB patients complicated by diabetes had higher risk to develop PMA. Diabetes has been associated with an increased risk of surgical site infections, higher hospitalization costs, postoperative blood transfusions, and prolonged hospital stays in patients underwent spinal surgery (16), which is consistent with our findings. Ugolini et al. found that diabetes can lead to increased postoperative blood loss in patients undergoing thoracic surgery because diabetes causes hypercoagulation and changes in microvessels that could negatively affect the retraction and vasoconstriction of the damaged microvessels before the hemostasis coagulation phase (17). Moreover, diabetes leads to an adverse effect in angiogenesis in healing wounds, which leads to a decrease in the density of blood vessels and capillaries (factors vital to wound healing) in the surgical incision, leading to prolonged wound healing time, increased postoperative incision drainage, and invisible blood loss, further aggravating the decrease in postoperative hemoglobin level (18–20). In summary, diabetes leads to increased postoperative blood loss and is the leading cause of PMA in patients with STB.
BMI was also found as an independent risk factor for PMA in STB patients. We found that a BMI of > 19.5 kg/m2 can reduce the risk of PMA. We referred to the BMI classification of the World Health Organization (WHO), as follows: normal (18.5 kg/m2 ≤ BMI < 25 kg/m2), overweight (25 kg/m2 ≤ BMI < 30 kg/m2), and obese (BMI ≥ 30 kg/m2) (21). BMI < 19.5 kg/m2 is considered nearly malnutrition, and BMI levels were negatively associated with the prevalence of anemia; thus, patients with STB with BMI < 19.5 kg/m2 have a higher risk of developing anemia (22). High BMI may be a favorable prognostic factor in patients with chronic diseases such as STB, as patients with a higher BMI may have higher reserves in nutrition, which can benefit the postoperative treatment of patients with STB (23, 24). However, overly high BMI is associated with adverse effects, as overweight patients have a higher incidence of low-back pain and spinal disease requiring multiple surgeries (25, 26). Obesity, defined as a BMI of ≥ 30 kg/m2, is closely related to longer hospitalization and greater intraoperative blood loss in patients undergoing spinal fusion surgery, indicating that an overly high BMI may cause greater perioperative blood loss in patients with STB (27). In our study, a BMI within 19.5–30 kg/m2 was determined to be beneficial in preventing PMA in patients with STB.
In our study, we found that the preoperative hemoglobin level is an independent risk factor for PMA in STB patients and whose value less than 120 g/L can raise the risk of PMA. Considering the WHO definition of anemia and our previous findings, we can conclude that preoperative anemia is an independent risk factor for PMA in STB patients (28). Anemia is closely associated with tuberculosis. The odds of developing tuberculosis in patients with anemia are 3.56 times higher than in patients without anemia, and 35% of patients undergoing orthopedic surgery have preoperative anemia (29, 30). Preoperative anemia is independently related to mortality, postoperative complications, and prolonged hospitalization in orthopedic surgeries (31). Seicean et al. reported that preoperative anemia should be considered as an independent risk factor for postoperative complications that warrants attention before an elective spine surgery, further emphasizing the importance of the preoperative hemoglobin level in predicting PMA in patients with STB (8). Interestingly, when anemia was complicated simultaneously with a known risk factor before operation, it led to a significant aggravation in the impact of this risk factor on the outcomes of noncardiac surgeries (32). Thus, preoperative anemia determined by the preoperative hemoglobin level had the greatest weight in the scoring scale.
Many studies have shown that surgical blood loss, operative trauma, and duration of anesthesia are risk factors for postoperative complications, including anemia (33, 34). In the current study, operation time, which can reflect surgical blood loss, operative trauma, and anesthesia duration, was identified as an independent risk factor for PMA. The average operation time in patients who presented with PMA was 220.13 ± 58.32 min, whereas that in patients who did not develop PMA was 196.08 ± 52.67 min (P = 0.002). The ROC curve analysis revealed that an operation time of 205 min was the optimal cutoff value for PMA. Controlling the operation time less than 205 min can decrease the possibility of PMA in patients with STB.
In STB focus debridement, anterior and posterior surgical approaches are usually performed. Our study found that the use of the posterior approach is an independent risk factor for PMA, whereas the use of the anterior approach is not. The anterior approach is superior to the posterior approach in terms of operation time, intraoperative blood loss, and correction of the Cobb angle (35, 36). Moreover, the advantages of anterior approach are also including adequate debridement, less muscle injury, better decompression, and better interbody fusion cage placement. However, due to the establishment of the surgical channel the anterior approach may lead to some unsatisfactory orthopedic outcome like more serious pain and even more complications in chest and abdomen (37). In contrast, the posterior surgical approach can not only achieve more steady implant fixation but also better kyphosis correction. However, it’s disadvantages including insufficient lesion removal and nerve decompression, increased surgical blood loss, and longer operation time (38). In summary, the anterior approach has better decompression effects and a lower PMA risk, whereas the posterior approach has better orthopedic effects but a higher PMA risk.
Patients with STB have poorer immunity, experience bigger operative trauma, and have a higher risk of developing postoperative adverse events (15, 39). The establishment of a scoring scale for predicting PMA after STB debridement surgery is crucial for the perioperative management of patients and can effectively help physicians in making decisions about postoperative treatment strategies. To our knowledge, this is the first study to quantify the risk of PMA in patients with STB. However, there were some limitations to the develop the scoring scale. Owing to the small sample size of the validation set, the diagnostic accuracy of the scale cannot be fully validated. As the sensitivity of the scale was not high, a significant rate of missed diagnosis may exist. In addition, there may have been other independent clinical risk factors that were not analyzed in the present study. Further studies are warranted to confirm the validity of the scoring scale and improve it for the risk assessment of PMA in patients with STB.
Studies on PMA in patients with STB are lacking. Accordingly, no study has provided recommendations on the adequate preoperative hemoglobin level in patients with STB. Therefore, our study may be the first of its kind. We found that one of the independent risk factors vital to the occurrence of PMA in patients with STB was the preoperative hemoglobin level. On the basis of the treatment principles of personalized and precision medicine, we believe that it is necessary to further explore the adequate preoperative hemoglobin level in patients with STB (40, 41). Therefore, we established a scoring scale for PMA in patients with STB, on the basis of which we divided the patients into the low-, moderate-, and high-risk groups. Thereafter, we obtained the values of the perioperative hemoglobin level changes in the three groups. By combining the WHO diagnostic criteria for moderate anemia (hemoglobin level < 90 g/L) and the values of the perioperative hemoglobin level changes in the three groups, the recommended preoperative hemoglobin levels for the three groups of patients with STB were determined as follows: low-risk group, 110 g/L; moderate-risk group, 115 g/L; and high-risk group, 120 g/L. Interestingly, the recommended preoperative hemoglobin level for the high-risk group was compatible with the diagnostic criteria for mild anemia, and the recommended preoperative hemoglobin levels for the low- and moderate-risk groups were both < 120 g/L, which indicates that preoperative mild anemia in some patients with STB is acceptable from the perspective of preventing PMA.
We found that the probability of PMA in low-risk patients is very low. For such patients, if the preoperative hemoglobin level is > 110 g/L, active intervention measures are not needed but close attention must be paid to postoperative hemoglobin changes. However, for the moderate- and high-risk groups, especially the high-risk group, if the preoperative hemoglobin level is lower than the recommended value, we believe that preoperative interventions are necessary, such as iron and erythropoietin supplementation for iron deficiency anemia, vitamin B12 supplementation for megaloblastic anemia, application of tranexamic acid before surgery, optimization of the surgical process, and preoperative blood transfusion if necessary (42, 43). We hope that the recommendations on the preoperative hemoglobin level in this study would help improve the trend in postoperative hemoglobin level changes and reduce the occurrence of PMA in patients with STB.
This study had some limitations. First, there may be some risk factors that we did not consider. Second, the sample size of the validation set was small. Third, we did not perform validation of the recommended preoperative hemoglobin levels. Future studies addressing these issues are required to confirm our results.
Five independent risk factors for PMA in STB patients were identified in this study and established a scoring scale. When score more than 4 points, it is considered to be associated with PMA development. This scale is determined by clinical data, making it suitable for use by physicians in predicting patient’s postoperative condition changes. According to the scoring scale, patients with STB with a score of 0–3 have a low risk of developing PMA; those with a score of 4–6 have a moderate risk; and those with a score of 7–10 have a high risk. The recommended preoperative hemoglobin levels are 110, 115, and 120 g/L in the low-, moderate-, and high-risk groups, respectively.
Authors’ Contributions
GJ and YO performed the conception and design; GJ and XD performed the data analysis and interpretation; GJ, XD, MZ and WQ performed the data collection and management; all authors performed the manuscript writing and critical revisions; GJ and YO did the overall responsibility. All authors read and approved the final manuscript.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Availability of data and materials
The clinical data of our study is accessible from the corresponding author on reasonable request.
Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by the Institutional Ethics Board of The First Affiliated Hospital of Chongqing Medical University. All study participants provided written informed consent to store their data in a hospital database and use it for research purposes.
Competing interests
The authors declare that they have no competing interests.
Table 1
Perioperative characteristics of 223 thoracic and lumbar tuberculosis patients.
|
Characteristics |
Postoperative moderate anemia |
P |
|
|
Yes (n=76) |
No (n=147) |
||
|
Age(year) |
52.26±16.81 |
44.84±15.46 |
0.001* |
|
BMI(kg/㎡) |
20.34±3.00 |
21.99±3.58 |
0.001* |
|
Sex(n,%) |
|
＜0.001* |
|
|
Female |
49 |
48 |
|
|
Male |
27 |
99 |
|
|
Diabetes mellitus |
|
＜0.001* |
|
|
Yes |
17 |
6 |
|
|
No |
59 |
141 |
|
|
Pulmonary tuberculosis |
|
0.938 |
|
|
Yes |
27 |
53 |
|
|
No |
49 |
94 |
|
|
Hepatitis |
|
0.804 |
|
|
Yes |
5 |
11 |
|
|
No |
71 |
136 |
|
|
Smoking history(year) |
6.64±13.28 |
8.47±12.85 |
0.327 |
|
Drinking history(year) |
4.29±10.76 |
4.90±10.42 |
0.683 |
|
Operative time(min) |
220.13±58.32 |
196.08±52.67 |
0.002* |
|
Operative blood loss(ml) |
436.96±332.87 |
317.29±234.71 |
0.002* |
|
Surgery approach |
|
0.047* |
|
|
Posterior approach |
75 |
135 |
|
|
Anterior-Posterior approach |
1 |
12 |
|
|
Number of diseased vertebrae |
2.36±1.31 |
2.11±0.85 |
0.141 |
|
Number of fixation segments |
4.96±1.22 |
4.48±1.29 |
0.007* |
|
Number of pedicle screws |
8.79±2.14 |
8.18±2.21 |
0.051 |
|
Preoperative hemoglobin(g/L) |
109.68±13.68 |
128.31±13.78 |
＜0.001* |
|
Preoperative lymphocytes(×109/L) |
1.15±0.40 |
1.30±0.50 |
0.015* |
|
Preoperative serum albumin(g/L) |
37.63±3.69 |
40.22±4.84 |
＜0.001* |
|
Preoperative CRP(mg/L) |
36.48±35.79 |
28.45±30.76 |
0.082 |
|
Preoperative ESR(mm/h) |
64.21±26.95 |
49.98±27.34 |
＜0.001* |
|
Course of disease(month) |
17.34±44.85 |
7.58±10.23 |
0.065 |
Abbreviations: BMI, body mass index; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.
Table 2
Univariate binary logistic regression analysis of postoperative moderate anemia.
|
Characteristics |
Regression coefficient(β) |
Odds ratio(OR) |
95% CI |
P |
|
Age |
0.029 |
1.030 |
1.011-1.049 |
0.001* |
|
Sex |
-1.320 |
0.267 |
0.149-0.478 |
＜0.001* |
|
BMI |
-0.153 |
0.858 |
0.783-0.941 |
0.001* |
|
Pulmonary tuberculosis |
0.023 |
1.023 |
0.574-1.824 |
0.938 |
|
Diabetes mellitus |
-1.913 |
0.148 |
0.055-0.393 |
＜0.001* |
|
Hepatitis |
0.138 |
1.149 |
0.384-3.434 |
0.804 |
|
Smoking history |
-0.011 |
0.989 |
0.967-1.011 |
0.321 |
|
Drinking history |
-0.006 |
0.994 |
0.968-1.022 |
0.682 |
|
Course of disease |
0.024 |
1.024 |
1.003-1.045 |
0.025* |
|
Number of diseased vertebrae |
0.232 |
1.261 |
0.935-1.700 |
0.129 |
|
Operation time |
0.008 |
1.008 |
1.003-1.013 |
0.003* |
|
Operation blood loss |
0.002 |
1.002 |
1.000-1.003 |
0.004* |
|
Surgery approach |
-1.897 |
0.150 |
0.019-1.176 |
0.047* |
|
Number of fixation segments |
0.297 |
1.346 |
1.072-1.689 |
0.010* |
|
Number of pedicle screws |
0.125 |
1.133 |
0.995-1.290 |
0.059 |
|
Preoperative hemoglobin |
-0.095 |
0.909 |
0.885-0.934 |
＜0.001* |
|
Preoperative lymphocytes |
-0.723 |
0.485 |
0.257-0.915 |
0.026* |
|
Preoperative serum albumin |
-0.148 |
0.863 |
0.801-0.929 |
＜0.001* |
|
Preoperative CRP |
0.007 |
1.007 |
0.999-1.016 |
0.091 |
|
Preoperative ESR |
0.019 |
1.019 |
1.008-1.030 |
＜0.001* |
Abbreviations: BMI, body mass index; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.
Table 3
Multivariate binary logistic regression analysis of postoperative moderate anemia.
|
Characteristics |
Regression coefficient(β) |
Crude odds ratio(OR) |
95% CI |
P |
|
Age |
0.014 |
1.014 |
0.987-1.041 |
0.302 |
|
Sex |
-0.830 |
0.436 |
0.185-1.030 |
0.058 |
|
BMI |
-0.125 |
0.882 |
0.781-0.986 |
0.029* |
|
Diabetes mellitus |
-2.475 |
0.084 |
0.021-0.337 |
＜0.001* |
|
Course of disease |
0.020 |
1.012 |
0.994-1.048 |
0.132 |
|
Operation time |
0.011 |
1.011 |
1.005-1.018 |
0.001* |
|
Operation blood loss |
0.001 |
1.001 |
1.000-1.002 |
0.164 |
|
Surgery approach |
-2.999 |
0.050 |
0.003-0.951 |
0.046* |
|
Number of fixation segments |
-0.162 |
0.851 |
0.574-1.261 |
0.421 |
|
Preoperative hemoglobin |
-0.093 |
0.911 |
0.884-0.938 |
＜0.001* |
|
Preoperative lymphocytes |
-0.473 |
0.623 |
0.273-1.422 |
0.261 |
|
Preoperative serum albumin |
0.010 |
1.010 |
0.926-1.102 |
0.826 |
|
Preoperative ESR |
0.012 |
1.012 |
0.997-1.026 |
0.119 |
Abbreviations: BMI, body mass index; ESR, erythrocyte sedimentation rate.
Table 4
Scoring scale for occurrence of postoperative moderate anemia.
|
Clinical characteristics |
Points |
|
BMI |
|
|
≥19.5 kg/m2 |
0 |
|
≤19.4 kg/m2 |
1 |
|
Diabetes mellitus |
|
|
No |
0 |
|
Yes |
3 |
|
Preoperative hemoglobin |
|
|
≥120 g/L |
0 |
|
≤119 g/L |
3 |
|
Surgery approach |
|
|
Anterior-Posterior approach |
0 |
|
Posterior approach |
1 |
|
Operation time |
|
|
≥205 minutes |
2 |
|
≤204 minutes |
0 |
Abbreviations: BMI, body mass index.
Table 5
Comparison of performance of the scoring scale on derivation set and validation set.
|
Groups |
Derivation set |
Validation set |
|||||
|
With PMA |
Without PMA |
Total |
With PMA |
Without PMA |
Total |
||
|
Clinical diagnosis |
With PMA |
65 |
44 |
109 |
24 |
17 |
41 |
|
Without PMA |
11 |
103 |
114 |
3 |
28 |
31 |
|
|
|
Total |
76 |
147 |
223 |
27 |
45 |
72 |
|
Sensitivity(%) |
59.6% |
58.5% |
|||||
|
Specificity(%) |
90.4% |
90.3% |
|||||
Abbreviations: PMA, postoperative moderate anemia.
Table 6
Incidence of PMA’s occurrence in three groups.
|
Groups |
Patients with PMA |
Total patients |
Incidence of PMA |
|
Low risk group (score≤3points group) |
11 |
114 |
9.7% |
|
Moderate risk group (4≤score≤6 points group) |
40 |
81 |
49.4% |
|
High risk group (score≥7 points group) |
25 |
28 |
89.3% |
Abbreviations: PMA, postoperative moderate anemia.
Table 7
Comparison of perioperative hemoglobin changing value divided by scoring scale in 3 groups.
|
Hemoglobin changing value (g/L) |
Groups |
P-value |
||||
|
Low risk group |
Moderate risk group |
High risk group |
P1 |
P2 |
P3 |
|
|
Preoperative hemoglobin |
123.06±9.62 |
112.07±14.31 |
105.36±9.80 |
＜0.001 |
＜0.001 |
0.023 |
|
Postoperative lowest hemoglobin |
105.91±12.92 |
88.77±11.59 |
76.79±10.29 |
＜0.001 |
＜0.001 |
0.002 |
|
Hemoglobin loss |
20.07±10.47 |
24.44±12.67 |
29.18±10.34 |
0.006 |
＜0.001 |
0.077 |
Footnote: P1: Comparison between low risk group and moderate risk group. P2: Comparison between low risk group and high risk group. P3: Comparison between moderate risk group and high risk group.
Table 8
Preoperative serum hemoglobin recommended value based on the scoring scale.
|
Groups |
Recommended value |
|
Low risk group (score≤3points group) |
110g/L |
|
Moderate risk group (4≤score≤6 points group) |
115g/L |
|
High risk group (score≥7 points group) |
120g/L |