After excluding subjects who did not the meet study criteria, a total of 91,016 patients received GA were included into the preliminary group for initial model training, and 590,763 patients were selected as the validation group for further analysis (Figure 1). The baseline characteristics of the preliminary group was presented in Table 1.
Table 1
Baseline characteristic of preliminary group
| All (n=91016) | Major organ transplantation (n=136) | CV (n=2060) | Neurosurgery (n=4851) | Others (n=83969) |
| | (%) | (%) | (%) | (%) | (%) |
Mean age, years (SD) | | 52.0 (17.2) | 49.6 (9.6) | 62.8 (14.6) | 56.9 (17.7) | 51.5 (17.1) |
Age | <65 | 67187 (73.8) | 134 (98.5) | 1032 (50.1) | 3033 (62.5) | 62988 (75.0) |
| 65-74 | 12977 (14.3) | 2 (1.5) | 534 (25.9) | 884 (18.2) | 11557 (13.8) |
| 75+ | 10852 (11.9) | 0 | 494 (24.0) | 934 (19.3) | 9424 (11.2) |
Gender | Female | 49359 (54.2) | 44 (32.4) | 697 (33.8) | 1829 (37.7) | 46789 (55.7) |
| Male | 41657 (45.8) | 92 (67.6) | 1363 (66.2) | 3022 (62.3) | 37180 (44.3) |
Surgery time | Within 2 hours | 44155 (48.5) | 0 | 142 (6.9) | 1535 (31.6) | 42478 (50.6) |
| 2-2.5 hr | 12014 (13.2) | 0 | 49 (2.4) | 601 (12.4) | 11364 (13.5) |
| 2.5-3 hr | 9301 (10.2) | 3 (2.2) | 53 (2.6) | 512 (10.6) | 8733 (10.4) |
| 3-3.5 hr | 6268 (6.9) | 10 (7.4) | 65 (3.2) | 382 (7.9) | 5811 (6.9) |
| 3.5-4 hr | 4869 (5.3) | 10 (7.4) | 99 (4.8) | 354 (7.3) | 4406 (5.2) |
| Over 4hr | 14409 (15.8) | 113 (83.1) | 1652 (80.2) | 1467 (30.2) | 11177 (13.3) |
Hospital level | center | 44702 (49.1) | 124 (91.2) | 1413 (68.6) | 2577 (53.1) | 40588 (48.3) |
| others | 46314 (50.9) | 12 (8.8) | 647 (31.4) | 2274 (46.9) | 43381 (51.7) |
Comorbidity: | | | | | | |
Congestive heart failure | | 3769 (4.1) | 33 (24.3) | 637 (30.9) | 242 (5.0) | 2857 (3.4) |
Cardiac arrhythmias | | 4650 (5.1) | 13 (9.6) | 440 (21.4) | 300 (6.2) | 3897 (4.6) |
Valvular disease | | 2552 (2.8) | 10 (7.4) | 489 (23.7) | 124 (2.6) | 1929 (2.3) |
Pulmonary circulation disorders | | 217 (0.2) | 6 (4.4) | 43 (2.1) | 11 (0.2) | 157 (0.2) |
Peripheral vascular disorders | | 1871 (2.1) | 2 (1.5) | 250 (12.1) | 125 (2.6) | 1494 (1.8) |
Hypertension | | 22440 (24.7) | 49 (36.0) | 1096 (53.2) | 1556 (32.1) | 19739 (23.5) |
Paralysis | | 958 (1.1) | 0 | 31 (1.5) | 122 (2.5) | 805 (1.0) |
Other neurological disorders | | 2106 (2.3) | 5 (3.7) | 52 (2.5) | 372 (7.7) | 1677 (2.0) |
Chronic pulmonary disease | | 10502 (11.5) | 18 (13.2) | 470 (22.8) | 591 (12.2) | 9423(11.2) |
Diabetes | | 11717 (12.9) | 30 (22.1) | 629 (30.5) | 699 (14.4) | 10359 (12.3) |
Hypothyroidism | | 2123 (2.3) | 0 | 26 (1.3) | 61 (1.3) | 2036 (2.4) |
Renal failure | | 3487 (3.8) | 60 (44.1) | 231 (11.2) | 201 (4.1) | 2995 (3.6) |
Liver disease | | 4175 (4.6) | 47 (34.6) | 84 (4.1) | 194 (4.0) | 3850 (4.6) |
Peptic ulcer disease excluding bleeding | | 12071 (13.3) | 31 (22.8) | 322 (15.6) | 509 (10.5) | 11209 (13.3) |
AIDS/HIV | | 41 (0.0) | 0 | 0 | 6 (0.1) | 35 (0.0) |
Lymphoma | | 221 (0.2) | 0 | 6 (0.3) | 24 (0.5) | 191 (0.2) |
Metastatic cancer | | 908 (1.0) | 0 | 24 (1.2) | 85 (1.8) | 799 (1.0) |
Solid tumor without metastasis | | 10389 (11.4) | 26 (19.1) | 151 (7.3) | 350 (7.2) | 9862 (11.7) |
Rheumatoid arthritis/ collagen vascular diseases | | 3895 (4.3) | 5 (3.7) | 91 (4.4) | 187 (3.9) | 3612 (4.3) |
Coagulopathy | | 406 (0.4) | 15 (11.0) | 21 (1.0) | 27 (0.6) | 343 (0.4) |
Obesity | | 505 (0.6) | 0 | 7 (0.3) | 21 (0.4) | 477 (0.6) |
Weight loss | | 699 (0.8) | 0 | 18 (0.9) | 44 (0.9) | 637 (0.8) |
Fluid and electrolyte disorders | | 2135 (2.3) | 9 (6.6) | 98 (4.8) | 160 (3.3) | 1868 (2.2) |
Blood loss anemia | | 877 (1.0) | 0 | 14 (0.7) | 16 (0.3) | 847 (1.0) |
Deficiency anemia | | 2074 (2.3) | 8 (5.9) | 42 (2.0) | 56 (1.2) | 1968 (2.3) |
Alcohol abuse | | 1306 (1.4) | 11 (8.1) | 18 (0.9) | 136 (2.8) | 1141 (1.4) |
Drug abuse | | 197 (0.2) | 0 | 6 (0.3) | 14 (0.3) | 177 (0.2) |
Psychoses | | 949 (1.0) | 0 | 16 (0.8) | 92 (1.9) | 841 (1.0) |
Depression | | 4021 (4.4) | 5 (3.7) | 66 (3.2) | 297 (6.1) | 3653 (4.4) |
Medication use: | | | | | | |
Biguanides | | 7679 (8.4) | 16 (11.8) | 407 (19.8) | 468 (9.6) | 6788 (8.1) |
Sulfonylurea | | 7968 (8.8) | 19 (14.0) | 453 (22.0) | 485 (10.0) | 7011 (8.3) |
Alpha-glucosidase inhibitors | | 1855 (2.0) | 6 (4.4) | 132 (6.4) | 108 (2.2) | 1609 (1.9) |
Thiazolidinediones | | 1885 (2.1) | 8 (5.9) | 110 (5.3) | 107 (2.2) | 1660 (2.0) |
Glinides | | 1414 (1.6) | 7 (5.1) | 105 (5.1) | 95 (2.0) | 1207 (1.4) |
Any oral anti-diabetic agents | | 10510 (11.5) | 28 (20.6) | 564 (27.4) | 630 (13.0) | 9288 (11.1) |
Fast-acting insulins | | 2938 (3.2) | 24 (17.6) | 245 (11.9) | 189 (3.9) | 2480 (3.0) |
ACE inhibitors | | 7989 (8.8) | 20 (14.7) | 710 (34.5) | 567 (11.7) | 6692 (8.0) |
Angiotensin receptor blockers | | 8027 (8.8) | 40 (29.4) | 647 (31.4) | 533 (11.0) | 6807 (8.1) |
Alpha-blockers | | 3243 (3.6) | 12 (8.8) | 184 (8.9) | 249 (5.1) | 2798 (3.3) |
Beta-blockers | | 18087 (19.9) | 68 (50.0) | 1122 (54.5) | 1166 (24.0) | 15731 (18.7) |
Calcium channel blockers | | 20624 (22.7) | 50 (36.8) | 1124 (54.6) | 1465 (30.2) | 17985 (21.4) |
Diuretics | | 13659 (15.0) | 81 (59.6) | 1010 (49.0) | 868 (17.9) | 11700 (13.9) |
Other anti-hypertensive agents | | 721 (0.8) | 14 (10.3) | 81 (3.9) | 41 (0.8) | 585 (0.7) |
Statins | | 7929 (8.7) | 21 (15.4) | 637 (30.9) | 421 (8.7) | 6850 (8.2) |
Fibrates | | 2879 (3.2) | 2 (1.5) | 156 (7.6) | 176 (3.6) | 2545 (3.0) |
Digitalis glycoside | | 1497 (1.6) | 20 (14.7) | 330 (16.0) | 92 (1.9) | 1055 (1.3) |
Anti-arrhythmics class I and III | | 1902 (2.1) | 18 (13.2) | 227 (11.0) | 145 (3.0) | 1512 (1.8) |
Mean no. of different prescription drugs (SD) | | 23.3 (17.2) | 34.0 (19.0) | 32.6 (20.8) | 22.7 (18.2) | 23.1 (17.0) |
Mean no. of cardiovascular-related medications (SD) | | 1.6 (2.7) | 4.9 (4.2) | 6.0 (4.6) | 2.0 (2.8) | 1.5 (2.5) |
Mean no. of outpatient visit (SD) | | 25.9 (22.4) | 37.4 (19.8) | 32.4 (25.7) | 24.9 (24.1) | 25.8 (22.2) |
Mean no. of emergency room visit (SD) | | 0.7 (2.1) | 2.1 (5.8) | 1.3 (2.1) | 0.9 (2.7) | 0.7 (2.0) |
Mean no. of hospitalization (SD) | | 0.3 (0.9) | 1.6 (1.7) | 0.9 (1.3) | 0.4 (1.0) | 0.3 (0.9) |
Mean days of hospitalization (SD) | | 2.5 (11.3) | 15.4 (23.8) | 6.1 (13.0) | 3.5 (13.1) | 2.4 (11.1) |
Mean follow-up days (SD) | | 29.8 (1.9) | 29.4 (2.8) | 29.0 (4.7) | 28.6 (5.4) | 29.9 (1.3) |
Mean person-year follow-up (SD) | | 0.0795 (0.0051) | 0.0785 (0.0074) | 0.0773 (0.0125) | 0.0764 (0.0143) | 0.0798 (0.0035) |
Sum of person-year follow-up | | 7239.48 | 10.68 | 159.17 | 370.8 | 6698.83 |
Resource utilization of index hospitalization | | | | | | |
days of hospitalization | Mean (SD) | 10.7 (31.6) | 41.2 (41.4) | 24.9 (37.7) | 28.2 (62.4) | 9.3 (28.2) |
| Median (Q1-Q3) | 5 (3-10) | 29 (15-51) | 17 (11-27) | 14 (7-31) | 5 (3-9) |
medical cost | Mean (SD) | 108187 (207469.9) | 1214102 (938757.1) | 484404 (384955.0) | 271852 (374019.7) | 87711 (161571.5) |
| Median (Q1-Q3) | 54891 (35950-102692) | 978193 (366156-1841393) | 428177 (323254-495202) | 171012 (80513-335195) | 53636 (35027-88189) |
Mortality Rate | | | | | | |
No. of death within 30 days (%) | | 962 (1.1) | 6 (4.4) | 125 (6.1) | 336 (6.9) | 495 (0.6) |
30 days mortality rate, per 100 py | (95% CI) | 12.94 (12.14-13.78) | 54.71 (24.58-121.78) | 76.48 (64.18-91.13) | 88.25 (79.3-98.21) | 7.19 (6.59-7.86) |
No. of death during admission (%) | | 1388 (1.5) | 11 (8.1) | 155 (7.5) | 384 (7.9) | 838 (1.0) |
Cardiovascular (CV), angiotensin-converting-enzyme inhibitor (ACE inhibitors), standard deviation (SD) |
Not surprisingly, antihypertensive agents were most frequently used in high-risk surgical patients, including angiotensin-converting-enzyme inhibitor, angiotensin receptor blockers, alpha-blockers, beta-blockers, calcium channel blockers, diuretics, and other anti-hypertensive agents in both groups.
Patients who underwent CV and major neurosurgical procedures had a higher crude incidence rate of death and mortality rate within 30 days (6.1 and 6.9%, separately). The medical cost was also associated with the duration of hospitalization after operation, and the patients who received major organ transplantation got the most medical resource utilization (Table 1).
In the crude analysis of factors associated with mortality in preliminary group, the major surgeries such as major organ transplantation, CV surgery, and neurosurgery, were associated with a significantly higher risk for mortality (Table 2). Among these, neurosurgery was associated with the highest risk of post-operation mortality with adjusted OR ranging from 6.94 to 10.37 (Table 2). After the stratified analysis, age also showed significant influence for determining post-operation mortality, especially in the elderly groups (Table 2).
Table 2
Factors associated with mortality in preliminary group by using logistic regression model with forward selection of covariates (p > 0.10 for exclusion, p < 0.05 for inclusion)
In-hospital mortality | 30-days mortality |
Covariates selected | Adjusted Odds Ratio | 95% CI | p-value | Covariates selected | Adjusted Odds Ratio | 95% CI | p-value |
Surgical types | | | <.0001 | Surgical types | | | <.0001 |
Others | Reference | | | Others | Reference | | |
CV | 4.25 | 3.49- 5.17 | | CV | 6.37 | 5.12- 7.92 | |
Neurosurgery | 6.94 | 6.08- 7.91 | | Neurosurgery | 10.37 | 8.94-12.03 | |
Organ transplant | 5.20 | 2.67-10.13 | | Organ transplant | 4.40 | 1.85-10.51 | |
Age | | | <.0001 | Age | | | <.0001 |
<65 years | Reference | | | <65 years | Reference | | |
65-74 years | 2.13 | 1.82- 2.49 | | 65-74 years | 1.96 | 1.63- 2.36 | |
>=75 years | 4.49 | 3.90- 5.16 | | >=75 years | 3.90 | 3.30- 4.60 | |
Female vs. Male | 0.65 | 0.58- 0.73 | <.0001 | Female vs. Male | 0.71 | 0.62- 0.82 | <.0001 |
Hospital level (others vs. medical center) | 1.13 | 1.01- 1.27 | 0.0307 | Hospital level (others vs. medical center) | 1.38 | 1.20- 1.57 | <.0001 |
Peripheral vascular disorders | 0.71 | 0.53-0.94 | 0.0186 | Other neurological disorders | 0.66 | 0.47-0.91 | 0.0102 |
Renal failure | 1.43 | 1.19- 1.71 | 0.0001 | Hypertension | 0.77 | 0.64- 0.93 | 0.0062 |
Liver disease | 1.25 | 1.01- 1.54 | 0.0367 | Renal failure | 1.59 | 1.27- 2.00 | <.0001 |
Weight loss | 1.77 | 1.23- 2.55 | 0.0023 | Liver disease | 1.44 | 1.12- 1.84 | 0.0043 |
Fluid and electrolyte disorders | 1.42 | 1.15- 1.75 | 0.0010 | Coagulopathy | 1.99 | 1.20- 3.28 | 0.0074 |
Blood loss anemia | 1.81 | 1.21- 2.69 | 0.0036 | Alcohol abuse | 2.43 | 1.74- 3.41 | <.0001 |
Alcohol abuse | 2.09 | 1.55- 2.82 | <.0001 | Psychoses | 1.89 | 1.22- 2.93 | 0.0044 |
Psychoses | 1.62 | 1.09- 2.39 | 0.0160 | Directics | 1.44 | 1.22- 1.69 | <.0001 |
Fast-acting insulins | 1.42 | 1.17- 1.71 | 0.0003 | Antiarrhythmics class I and III | 1.45 | 1.11- 1.89 | 0.0060 |
Diuretics | 1.53 | 1.34- 1.75 | <.0001 | no. of emergency room visit | 1.02 | 1.00- 1.03 | 0.0189 |
Statins | 0.80 | 0.67- 0.96 | 0.0133 | no. of hospitalization | 1.06 | 1.01- 1.12 | 0.0217 |
Anti-arrhythmics class I and III | 1.51 | 1.22- 1.86 | 0.0002 | no. of outpatient visit | 0.99 | 0.99-1.00 | 0.0005 |
no. of outpatient visit | 0.99 | 0.99-1.00 | <0.0001 | | | | |
no. of emergency room visit | 1.01 | 1.00- 1.03 | 0.0480 | | | | |
no. of hospitalization | 1.10 | 1.05- 1.15 | <.0001 | | | | |
Cardiovascular (CV), 95% confidence interval (95% CI) |
For subgroup analysis, variables of the underlying comorbidities were almost positively related to post-operation mortality (Table 2). Unexpectedly, hypertension was not associated with increasing risk of post-operation mortality (adjusted OR: 0.77, CI: 0.64- 0.93) as other comorbidities were (Table 2). Nevertheless, prescription with statins usage seemed to be protective for the patient accepted GA (adjusted OR: 0.8, CI: 0.67- 0.96, Table 2).
The results of factors associated with post-operation mortality for different surgery types in preliminary group were showed in Table 3 (in-hospital mortality and 30-days mortality). As different surgical types were considered, age was still a significantly determining factor associated with post-operation mortality (Table 3).
Table 3
Factors associated with in-hospital and 30-day mortality for different surgery types in preliminary group
CV | | | Neurosurgery | | | Transplantation | | | Other | | |
In-hospital mortality |
Covariates selected | AOR | 95% CI | Covariates selected | AOR | 95% CI | Covariates selected | AOR | 95% CI | Covariates selected | AOR | 95% CI |
Area under ROC curves | 0.6652 | | Area under ROC curves | 0.6467 | | Area under ROC curves | 0.8418 | | Area under ROC curves | 0.8677 | |
Age | | | Age | | | Others vs. medical center | 8.91 | 1.41-56.17 | Age | | |
<65 y/o | ref. | | <65 y/o | ref. | | Alpha-glucosidase inhibitors | 9.45 | 0.68-130.6 | <65 y/o | ref. | |
65-74 y/o | 1.49 | 0.97- 2.28 | 65-74 y/o | 1.22 | 0.92- 1.62 | Glinides | 14.15 | 1.46-136.8 | 65-74 y/o | 2.99 | 2.42- 3.69 |
>=75 y/o | 2.10 | 1.41- 3.13 | >=75 y/o | 1.62 | 1.25- 2.10 | Fast-acting insulins | 4.31 | 0.93-19.97 | >=75 y/o | 7.40 | 6.15- 8.90 |
Valvular disease | 0.45 | 0.28- 0.72 | Others vs. medical center | 1.50 | 1.21- 1.86 | no. of outpatient visit | 0.92 | 0.87- 0.98 | Female vs. Male | 0.52 | 0.45- 0.61 |
Hypertension | 0.63 | 0.44- 0.90 | Cardiac arrhythmias | 1.51 | 1.05- 2.17 | | | | Valvular disease | 1.43 | 1.07- 1.91 |
Lymphoma | 5.51 | 0.95-32.15 | Diabetes, uncomplicated | 1.61 | 1.23- 2.11 | | | | Peripheral vascular disorders | 0.45 | 0.29-0.70 |
Beta-blockers | 0.65 | 0.45- 0.93 | Renal failure | 2.27 | 1.54- 3.34 | | | | Hypertension | 0.80 | 0.66- 0.96 |
Statins | 0.58 | 0.38- 0.89 | AIDS/HIV | 5.62 | 0.98-32.30 | | | | Renal failure | 1.27 | 1.02- 1.59 |
no. of emergency room visit | 1.07 | 1.00- 1.16 | Lymphoma | 3.05 | 1.09- 8.54 | | | | Solid tumor without metastasis | 0.74 | 0.60- 0.92 |
no. of hospitalization | 1.14 | 1.01- 1.30 | Coagulopathy | 3.89 | 1.65- 9.17 | | | | Weight loss | 1.91 | 1.27- 2.88 |
| | | Other neurological disorders | 0.33 | 0.18-0.60 | | | | Fluid and electrolyte disorders | 1.40 | 1.10- 1.79 |
| | | | | | | | | Blood loss anemia | 1.72 | 1.11- 2.68 |
| | | | | | | | | Alcohol abuse | 2.79 | 1.96- 3.97 |
| | | | | | | | | Psychoses | 2.38 | 1.52- 3.72 |
| | | | | | | | | Thiazolidinediones | 0.63 | 0.41- 0.96 |
| | | | | | | | | Fast-acting insulins | 1.38 | 1.10- 1.74 |
| | | | | | | | | Directics | 1.85 | 1.56- 2.19 |
| | | | | | | | | Antiarrhythmics class I and III | 1.68 | 1.29- 2.18 |
| | | | | | | | | no. of outpatient visit | 0.99 | 0.99- 1.00 |
| | | | | | | | | no. of emergency room visit | 1.02 | 1.01- 1.03 |
| | | | | | | | | no. of hospitalization | 1.10 | 1.05- 1.16 |
30-day mortality |
Covariates selected | AOR | 95% CI | Covariates selected | AOR | 95% CI | Covariates selected | AOR | 95% CI | Covariates selected | AOR | 95% CI |
Age | | | Age | | | Pulmonary circulation disorders | >999.999 | - | Age | | |
<65 yy/o | ref. | | <65 y/o | ref. | | Solid tumor without metastasis | >999.999 | - | <65 y/o | ref. | |
65-74 y/o | 1.15 | 0.70- 1.88 | 65-74 y/o | 1.18 | 0.87- 1.59 | no. of outpatient visit | 0.15 | <0.0001-141.07 | 65-74 y/o | 3.18 | 2.44- 4.14 |
>=75 y/o | 2.07 | 1.34- 3.18 | >=75 y/o | 1.53 | 1.16- 2.01 | Hypertension, uncomplicated | >999.999 | - | >=75 y/o | 7.33 | 5.78- 9.29 |
Valvular disease | 0.39 | 0.22- 0.69 | Others vs. medical center | 1.81 | 1.44- 2.27 | Antiarrhythmics class I and III | 0.004 | - | Female vs. Male | 0.57 | 0.47- 0.68 |
Hypertension | 0.54 | 0.36- 0.81 | Cardiac arrhythmias | 1.64 | 1.12- 2.39 | | | | Others vs. medical center | 1.23 | 1.02- 1.48 |
Lymphoma | 8.89 | 1.49-52.96 | Renal failure | 2.32 | 1.55- 3.50 | | | | Valvular disease | 1.75 | 1.23- 2.50 |
Psychoses | 4.09 | 1.09-15.39 | Coagulopathy | 4.21 | 1.72-10.27 | | | | Peripheral vascular disorders | 0.37 | 0.19-0.70 |
Beta-blockers | 0.56 | 0.38- 0.83 | Other neurological disorders | 0.37 | 0.20-0.69 | | | | Hypertension | 0.65 | 0.50- 0.83 |
no. of emergency room visit | 1.09 | 1.01- 1.17 | | | | | | | Renal failure | 1.44 | 1.08- 1.92 |
| | | | | | | | | Liver disease | 1.38 | 1.01- 1.87 |
| | | | | | | | | Solid tumor without metastasis | 0.76 | 0.57- 1.00 |
| | | | | | | | | Alcohol abuse | 3.85 | 2.55- 5.80 |
| | | | | | | | | Psychoses | 2.82 | 1.69- 4.71 |
| | | | | | | | | Directics | 1.78 | 1.44- 2.20 |
| | | | | | | | | Antiarrhythmics class I and III | 1.46 | 1.02- 2.08 |
| | | | | | | | | no. of outpatient visit | 0.99 | 0.99- 1.00 |
| | | | | | | | | no. of emergency room visit | 1.02 | 1.00- 1.03 |
| | | | | | | | | no. of hospitalization | 1.09 | 1.02- 1.15 |
Cardiovascular (CV), Adjusted Odds Ratio (AOR), 95% confidence interval (95% CI), Receiver operating characteristic curve (ROC) |
For patients accepting CV surgery, the comorbidity of “valvular heart disease”, or “hypertension” was not associated with increasing risk of post-operation mortality (Table 3). Medication with beta-blockers or statins usage seemed to be protective for the patients undergoing CV operation (Table 3). For the neurosurgery, age and underlying comorbidities were important covariates for the post-operation mortality (Table 3). As major organ transplantation was considered, the hospital level and the underlying comorbidities might determine the survival rate of the patients after operation. Unsurprisingly, patients accepted organ transplantation at the medical center got better outcomes (Table 3).
For the less risk surgeries, increasing age was associated with a significantly higher risk for post-operation mortality. Interesting, hypertension and prescription of thiazolidinediones (TZD) for diabetes mellitus (DM) were with decreasing post-operation mortality (Table 3) in these less risk surgeries. We also observed “valvular heart disease” was associated with increasing post-operation mortality among less risk surgeries, and that phenomenon was converse among the category of CV surgery (Table 3).
The prediction model was established according to the result of logistic regression model with forward selection of covariates and formulated as coefficient of each risk factor as e-Table 5 and e-Table 6 in the supplement showed. By applying individual’s parameters into the formula, it can be used to estimate the probability of 30 –day mortality and in-hospital mortality of each patient.
The ROC analysis of this prediction model was built in preliminary group initially. (Figure 2A and 2B) The optimal AUROC generated from the largest sensitivity and specificity summation for the prediction model of mortality after operation under GA was 0.8746 for predicating in-hospital mortality (Figure 2A) and 0.8698 for 30-days mortality (Figure 2B). After the prediction model was built from the preliminary group, further evaluating of the accuracy was conducted in the validation group. The prediction model based on the preliminary group also performed well in the validation group for post-operation mortality. (AUROC=0.8753 for in-hospital mortality; AUROC= 0.8767 for 30-days mortality) (Figure 2C and 2D). Validation of this prediction model demonstrated a high level of sensitivity and reliability.