3.3.1 Blood pressure and heart rate indicators
Among the 1,511 consultation records, 164 patients were assessed as high-risk by the internist and required medication intervention. After temporary oral medication control, these surgeries were completed. Two cases were deemed extremely high-risk and were terminated. Thus, 166 (10.52%) consultation records were included in the medication intervention group for analysis.
Of the patients who received intraoperative medication, 83.73% used nitroglycerin, with doses ranging from 0.3 to 2 mg. A few patients used captopril, metoprolol, or a combination of nitroglycerin and captopril, depending on their condition. A detailed summary is available in Table 4.
Table 4
Intervention during surgery. N = 166.
Type | Number |
Nitroglycerin | 139(83.73%) |
Captopril | 10(6.02%) |
Captopril + Nitroglycerin | 8(4.82%) |
Metoprolol | 7(4.22%) |
Surgical abortion | 2(1.20%) |
Total | 166(100%) |
Univariate analysis indicated that patient gender and various comorbidities, as well as whether a cardiac stent was implanted, influenced the need for intraoperative medication intervention. Female patients (t=-2.364, P = 0.019), patients without coronary heart disease (t = 3.617, P < 0.001), without arrhythmias (t = 2.750, P = 0.006), and without cerebrovascular disease (t = 2.469, P = 0.014) were more likely to need intraoperative blood pressure intervention. Patients without a cardiac stent (t = 2.562, P = 0.011), not taking anticoagulants (t = 4.439, P < 0.001), and not on statins (t = 4.477, P < 0.001) were also more likely to need intraoperative blood pressure intervention. Higher systolic blood pressure (t=-23.895, P < 0.001) and heart rate (t=-3.667, P < 0.001) preoperatively, and higher systolic blood pressure (t=-13.203, P < 0.001) and heart rate (t=-3.496, P < 0.001) at the start of the operation were associated with a greater likelihood of requiring intraoperative blood pressure intervention. Patients needing surgical suturing (t = 2.529, P = 0.012) were also more likely to require intraoperative blood pressure intervention. Refer to Table 5 for detailed information.
Table 5
Univariate analysis of the intervention group. N = 1511.
| No intervention group, n = 1345 | The intervention group, n = 166 | t | P-value |
Sex | | |
Male (Ref) | 655(90.97%) | 65(9.03%) | -2.364 | 0.0 19* |
Female | 690(87.23%) | 101(12.77%) | |
Age | 73.70 ± 11.57 | 74.01 ± 12.75 | -0.326 | 0.745 |
Coronary heart disease | | | | |
Yes | 460(92.74%) | 36(7.26%) | 3.617 | <0.001*** |
No | 885(87.19%) | 130(12.81%) | |
Arrhythmia | | |
Yes | 264(92.96%) | 20(7.04%) | 2.750 | 0.006** |
No | 1081(88.10%) | 146(11.90%) | |
Heart stent | | | | |
Yes | 316(92.40%) | 26(7.60%) | 2.562 | 0.011* |
No | 1029(88.02%) | 140(11.98%) | | |
Anticoagulant | | | | |
Yes | 827(91.99%) | 72(8.01%) | 4.439 | <0.001*** |
No | 518(84.64%) | 94(15.36%) | | |
Statins | | | | |
Yes | 626(92.74%) | 49(7.26%) | 4.477 | <0.001*** |
No | 719(86.00%) | 117(14.00%) | | |
Cerebrovascular disease | | | | |
Yes | 283(92.48%) | 23(7.52%) | 2.469 | 0.014* |
No | 1062(88.13%) | 143(11.87%) | | |
Diabetes | | | | |
Yes | 424(89.83%) | 48(10.17%) | 0.684 | 0.494 |
No | 921(88.64%) | 118(11.36%) | | |
Preoperative SBP | 137.97 ± 12.21 | 167.78 ± 14.90 | -23.895 | <0.001*** |
Preoperative HR | 74.82 ± 12.42 | 79.27 ± 15.02 | -3.667 | <0.001*** |
At the start of the operation’s SBP | 140.74 ± 16.10 | 158.22 ± 16.03 | -13.203 | <0.001*** |
At the start of the operation’s HR | 76.48 ± 12.54 | 80.87 ± 15.58 | -3.496 | <0.001*** |
Number of tooth extractions per time | 1.51 ± 1.00 | 1.57 ± 1.38 | -0.579 | 0.564 |
Dental high-speed turbine handpieces |
Yes | 34(85.00%) | 6(15.00%) | -0.822 | 0.411 |
No | 1311(89.12%) | 160(10.88%) | | |
Surgical incision | | | | |
Yes | 155(85.64% | 26(14.36%) | -1.398 | 0.164 |
No | 1190(89.47%) | 140(10.53%) | | |
Surgical suture | | | | |
Yes | 199(83.61%) | 39(16.39%) | 2.529 | 0.012* |
No | 1146(90.02%) | 127(9.98%) | | |
To avoid missing potential influencing factors, all variables from the univariate analysis were included in the logistic regression model. After controlling for confounding factors, it was found that higher systolic blood pressure preoperatively (OR = 1.146, P < 0.001), higher systolic blood pressure (OR = 1.026, P = 0.005), and higher heart rate (OR = 1.068, P = 0.002) at the start of the operation increased the likelihood of requiring medication intervention. The need for suturing also indicated a higher likelihood of requiring intraoperative medication intervention (OR = 2.271, P = 0.034). Comorbidities such as heart disease, cerebrovascular disease, and diabetes were not statistically significantly related to the need for intraoperative medication intervention. Detailed information is presented in Table 6.
Table 6
Multivariate analysis of the blood pressure intervention group. N = 1511.
| OR | OR 95%Low | OR 95%High | P-value |
Sex | 1.097 | 0.696 | 1.729 | 0.689 |
Age | 0.982 | 0.962 | 1.002 | 0.073 |
Coronary heart disease | 0.772 | 0.343 | 1.740 | 0.532 |
Arrhythmia | 1.177 | 0.636 | 2.179 | 0.605 |
Heart stent | 0.858 | 0.370 | 1.990 | 0.721 |
Anticoagulant | 1.068 | 0.624 | 1.828 | 0.811 |
Statins | 1.291 | 0.693 | 2.406 | 0.420 |
Cerebrovascular disease | 1.328 | 0.733 | 2.405 | 0.350 |
Diabetes | 0.814 | 0.510 | 1.299 | 0.388 |
Preoperative SBP | 1.146 | 1.122 | 1.170 | <0.001*** |
Preoperative HR | 0.962 | 0.922 | 1.003 | 0.069 |
At the start of the operation’s SBP | 1.026 | 1.008 | 1.044 | 0.005** |
At the start of the operation’s HR | 1.068 | 1.024 | 1.113 | 0.002** |
Number of tooth extractions per time | 1.024 | 0.837 | 1.252 | 0.820 |
Dental high-speed turbine handpieces | 0.732 | 0.215 | 2.487 | 0.617 |
Surgical incision | 1.341 | 0.542 | 3.317 | 0.525 |
Surgical suture | 2.271 | 1.064 | 4.848 | 0.034* |
ROC curves were plotted for the continuous variables identified. Systolic blood pressure preoperatively, systolic blood pressure at the start of the operation, and heart rate at the start of the operation could estimate the need for intraoperative medication intervention. The first measured systolic blood pressure preoperatively had the highest predictive efficiency. Using 158.5 mmHg as the threshold for systolic blood pressure preoperatively yielded the highest Youden index, at 0.69. With this cutoff, the sensitivity was 79.52%, and the specificity was 89.67%. Using 155.5 mmHg as the threshold for systolic blood pressure at the start of the operation yielded a Youden index of 0.43. With this cutoff, the sensitivity was 59.64%, and the specificity was 78.54%. For detailed information on the ROC curves, see Table 7 and Fig. 4.
Table 7
The area under the curve of preoperative SBP, at the start of the operation’s SBP and HR.
| The area under the curve | P value |
Preoperative SBP | 0.91 | <0.001*** |
At the start of the operation’s SBP | 0.79 | <0.001*** |
At the start of the operation’s HR | 0.58 | <0.001*** |
The sensitivity and specificity of these indicators were calculated based on the optimal ROC curve thresholds (AUC > 0.60). Using only systolic blood pressure preoperatively as a judgment indicator, the sensitivity was 79.52%, and the specificity was 89.67%. Using only systolic blood pressure at the start of the operation, the sensitivity was 59.64%, and the specificity was 78.54%. The indicator of surgical suturing had reasonable specificity (85.20%) but poor sensitivity (23.49%). When both systolic blood pressure preoperatively and at the start of the operation were used in combination, the combined specificity reached 94.28%, and the combined sensitivity was 87.95%, with a specificity of 79.18% and a Youden index of 0.67. The summary of the sensitivity and specificity of each indicator is presented in Table 8.
Table 8
Evaluation of SBP (preoperative and at the start of operation) and the need for suturing in predicting intraoperative intervention
| Sensitivity | Specificity | Positive Predictive Value | Negative Predictive Value | Youden’s J Statistic |
Preoperative SBP(158.5mmHg) | 79.52% | 89.67% | 48.71% | 97.26% | 0.69 |
At the start of the operation’s SBP(155.5mmHg) | 59.64% | 78.54% | 31.23% | 94.39% | 0.38 |
Surgical suture | 23.49% | 85.20% | 16.39% | 90.02% | 0.09 |
Preoperative SBP and at the start of the operation’s SBP | 51.20% | 94.28% | 49.63% | 93.93% | 0.45 |
Preoperative SBP or at the start of the operation’s SBP | 87.95% | 79.18% | 29.58% | 98.07% | 0.67 |
Preoperative SBP and surgical suture | 18.07% | 98.59% | 61.22% | 90.70% | 0.17 |
Preoperative SBP or surgical suture | 84.94% | 76.28% | 30.65% | 97.62% | 0.61 |
At the start of the operation’s SBP and surgical suture | 13.25% | 97.47% | 39.29% | 90.10% | 0.11 |
At the start of the operation’s SBP or surgical suture | 69.88% | 71.52% | 23.25% | 95.06% | 0.41 |
In clinical applications, sensitivity should be prioritized to ensure patient safety. Referring to the Youden index, a combination of systolic blood pressure preoperatively and at the start of the operation should be used to identify high-risk patients early. Blood pressure should be measured once preoperatively and once before the procedure begins. Patients with systolic blood pressure ≥ 159 mmHg preoperatively or ≥ 156 mmHg at the start of the operation should be prepared with antihypertensive medication and closely monitored for intraoperative blood pressure changes. In our hospital’s analyzed population, patients screened using this method had relatively safe treatments, with 98.07% maintaining stable intraoperative blood pressure without requiring temporary oral medication intervention.
3.3.2 ECG abnormality indicators
Among the 1,511 consultation records in this study, 625 cases (41.36%) showed new ECG abnormalities detected by continuous monitoring during surgery. The most common abnormalities were ventricular and supraventricular premature contractions, accounting for 97.76% of the new ECG findings. Only 25 cases involved frequent premature contractions (13 ventricular and 12 supraventricular), with 6 cases exhibiting both ventricular and supraventricular premature contractions, and 1 case showing both ventricular premature contractions and supraventricular tachycardia. The summary of abnormal ECG findings is presented in Table 9.
Table 9
Types of abnormal cardiac rhythms
Type N = 625 | n(%) |
Ventricular premature contraction | 335(53.60%) |
-Frequent premature ventricular contractions | 13(2.08%) |
Premature atrial contraction | 245 (39.20%) |
-Frequent premature atrial contraction | 12(1.92%) |
Ventricular premature contraction + Premature atrial contraction | 6(0.96%) |
Sinus arrhythmia | 7(1.12%) |
Atrioventricular conduction block | 4(0.64%) |
Supraventricular tachycardia | 2(0.32%) |
Ventricular premature contraction + Supraventricular tachycardia | 1(0.16%) |
Univariate analysis indicated that patient age, presence of coronary heart disease, cerebrovascular disease, use of anticoagulants, use of statins, preoperative systolic blood pressure, and systolic blood pressure at the start of the operation were statistically correlated with the occurrence of new intraoperative ECG abnormalities. The results of the univariate analysis are presented in Table 10. Patients without coronary heart disease (t = 2.262, P = 0.024), without cerebrovascular disease (t = 2.323, P = 0.020), not using anticoagulants (t = 2.107, P = 0.035), and not using statins (t = 2.660, P = 0.008) were more likely to experience abnormal ECG results during surgery. Additionally, older patients (t=-5.580, P < 0.001) and those with higher preoperative systolic blood pressure (t=-9.712, P < 0.001) and systolic blood pressure at the start of the operation (t=-8.631, P < 0.001) were more likely to experience intraoperative ECG abnormalities.
Table 10
Univariate analysis of abnormal electrocardiogram group. N = 1511
| Normal Electrocardiogram group | Abnormal electrocardiogram group | t | P-value |
Sex | | |
Male | 428(59.44%) | 292(40.56%) | -0.608 | 0.543 |
Female | 458(57.90%) | 333(42.10%) | |
Age | 72.34 ± 11.97 | 75.69 ± 11.04 | -5.580 | <0.001*** |
Coronary heart disease |
Yes | 311(62.70%) | 185(37.30%) | 2.262 | 0.024* |
No | 575(56.65%) | 440(43.35%) | |
Arrhythmia | | |
Yes | 181(63.73%) | 103(36.27%) | 1.964 | 0.050 |
No | 705(57.46%) | 522(42.54%) | |
Heart stent | | | | |
Yes | 215(62.87%) | 127(37.13%) | 1.826 | 0.068 |
No | 671(57.40%) | 498(42.60%) | | |
Anticoagulant | | | | |
Yes | 547(60.85%) | 352(39.15%) | 2.107 | 0.035* |
No | 339(55.39%) | 273(44.61%) | | |
Statins | | | | |
Yes | 421(62.37%) | 254(37.63%) | 2.660 | 0.008** |
No | 465(55.62%) | 371(44.38%) | | |
Cerebrovascular disease | |
Yes | 197(64.38%) | 109(35.62%) | 2.323 | 0.020* |
No | 689(57.18%) | 516(42.82%) | | |
Diabetes | | | | |
Yes | 294(62.29%) | 178(37.71%) | 1.958 | 0.050 |
No | 592(56.98%) | 447(43.02%) | | |
PreoperativeSBP | 137.22 ± 17.71 | 146.95 ± 20.17 | -9.712 | <0.001*** |
Preoperative HR | 75.51 ± 13.22 | 75.02 ± 12.19 | 0.732 | 0.464 |
At the start of the operation’s SBP | 139.57 ± 16.29 | 147.05 ± 17.01 | -8.631 | <0.001*** |
At the start of the operation’s HR | 77.13 ± 13.34 | 76.71 ± 12.45 | 0.622 | 0.534 |
Number of tooth extractions per time | 1.50 ± 0.99 | 1.56 ± 1.13 | -1.068 | 0.286 |
Dental high-speed turbine handpieces | |
Yes | 22(55.00%) | 18(45.00%) | -0.437 | 0.636 |
No | 864(58.42%) | 607(41.26%) | | |
Surgical incision |
Yes | 109(60.22%) | 72(39.78%) | 0.461 | 0.645 |
No | 777(58.42%) | 553(41.58%) | | |
Surgical suture |
Yes | 143(60.08%) | 95(39.92%) | -0.494 | 0.622 |
No | 743(58.37%) | 530(41.63%) | | |
All variables from the univariate analysis were included in the logistic regression model. The results of the multivariate analysis are summarized in Table 11. Patient age (OR = 1.033, P < 0.001), preoperative arrhythmia (OR = 1.020, P = 0.017), presence of cerebrovascular disease (OR = 1.490, P = 0.005), and preoperative systolic blood pressure (OR = 1.021, P < 0.001) were statistically correlated with new intraoperative ECG abnormalities.
Table 11
Multivariate analysis of abnormal electrocardiogram group
| OR | OR 95%Low | OR 95%High | P-value |
Sex | 1.163 | 0.927 | 1.458 | 0.192 |
Age | 1.022 | 1.012 | 1.033 | <0.001*** |
Coronary heart disease | 1.151 | 0.784 | 1.689 | 0.472 |
Arrhythmia | 1.436 | 1.068 | 1.931 | 0.017** |
Heart stent | 1.020 | 0.696 | 1.497 | 0.918 |
Anticoagulant | 0.894 | 0.681 | 1.172 | 0.417 |
Statins | 1.114 | 0.831 | 1.493 | 0.472 |
Diabetes | 1.146 | 0.906 | 1.450 | 0.255 |
Cerebrovascular disease | 1.490 | 1.129 | 1.967 | 0.005** |
Preoperative SBP | 1.021 | 1.012 | 1.030 | <0.001*** |
Preoperative HR | 0.991 | 0.968 | 1.013 | 0.415 |
At the start of the operation’s SBP | 1.008 | 0.998 | 1.018 | 0.111 |
At the start of the operation’s HR | 1.003 | 0.981 | 1.026 | 0.763 |
Number of tooth extractions per time | 1.031 | 0.926 | 1.147 | 0.581 |
Dental high-speed turbine handpieces | 0.772 | 0.390 | 1.531 | 0.459 |
Surgical incision | 1.042 | 0.637 | 1.704 | 0.869 |
Surgical suture | 1.053 | 0.690 | 1.606 | 0.812 |
ROC curves were plotted for the continuous variables identified, with preoperative systolic blood pressure remaining the most efficient predictor. For more details on the ROC curves, please refer to Table 12 and Fig. 5. Patient age also showed some predictive significance. The areas under the curves for preoperative systolic blood pressure and age were 0.64 and 0.58, respectively, indicating low diagnostic accuracy. The highest Youden index for preoperative systolic blood pressure was 0.20 at 145.5 mmHg, with a sensitivity of 53.60% and specificity of 66.82%. With the previous cutoff of 158.5 mmHg for intraoperative intervention, the Youden index was 0.19, with a sensitivity of 28.96% and a specificity of 89.84%. Compared to the 145.5 mmHg cutoff, the 158.5 mmHg threshold had significantly higher specificity and better-screened patients with relatively stable treatment processes. Therefore, the 158.5 mmHg threshold was adopted for diagnosis in this study.
Table 12
The area under the curve of preoperative SBP and age
| The area under the curve | P value |
Preoperative SBP | 0.64 | <0.001*** |
Age | 0.58 | <0.001*** |
Using preoperative systolic blood pressure with a threshold of 158.5 mmHg and adding the indicators of cerebrovascular disease and preoperative arrhythmia, the sensitivity, specificity, positive predictive value, negative predictive value, and Youden index were calculated. The results are presented in Table 13. These indicators showed slightly lower sensitivity but higher specificity. Notably, using preoperative systolic blood pressure as the standard, the sensitivity was 28.96%, and the specificity was 89.84%. When preoperative systolic blood pressure was below 158.5 mmHg, there was a 64.19% probability of no new intraoperative ECG abnormalities. Above this value, there was a 66.79% probability of new intraoperative ECG abnormalities. Referring to the Youden index, preoperative systolic blood pressure alone was the standard that had the highest Youden index of 0.19 for identifying patients likely to develop ECG abnormalities.
Table 13
Assessment of preoperative SBP, SBP at the start of the operation, and the need for suturing in predicting abnormal intraoperative ECGs
| Sensitivity | Specificity | Positive Predictive Value | Negative Predictive Value | Youden’s J Statistic |
Preoperative SBP(158.5mmHg) | 28.96% | 89.84% | 66.79% | 64.19% | 0.19 |
Arrhythmia | 16.48% | 79.57% | 35.71% | 57.43% | -0.04 |
Cerebrovascular disease | 17.44% | 77.77% | 35.62% | 57.18% | -0.05 |
Preoperative SBP(158.5mmHg)or arrhythmia | 40.32% | 72.01% | 50.40% | 63.11% | 0.12 |
Preoperative SBP(158.5mmHg)and arrhythmia | 3.84% | 98.53% | 64.86% | 59.23% | 0.02 |
Cerebrovascular disease or preoperative SBP(158.5mmHg) | 42.24% | 69.53% | 49.44% | 63.05% | 0.12 |
Cerebrovascular disease and preoperative SBP(158.5mmHg) | 4.16% | 98.08% | 60.47% | 40.80% | 0.02 |
Cerebrovascular disease or arrhythmia | 29.92% | 62.98% | 36.31% | 56.02% | -0.07 |
Cerebrovascular disease and arrhythmia | 2.72% | 95.49% | 29.82% | 58.18% | -0.02 |