Document resource
Introduction This study sought to characterise the in-hospital care of patients who suffered a type-one perioperative myocardial infarction (PMI) after orthopaedic, vascular, urological and gastrointestinal surgeries, and investigate the short and long-term all-cause mortality burden.Methods We analysed data from 472,278 adults from England and Wales hospitalised with AMI between Jan 2010 and March 2019 from the Myocardial Ischaemia National Audit Project (MINAP) registry, linked with Office for National Statistics mortality data and Hospital Episode Statistics, from where diagnosis of perioperative myocardial infarction was obtained from relevant International Classification of Diseases (ICD)-10 and Operative Procedure Codes Supplement (OPCS4) codes. Presence of a MINAP record confirms a final clinical diagnosis of type-one MI, rather than type-two MI.Patients admitted with a primary diagnosis of AMI that underwent inpatient surgery were excluded, including only PMI patients with a non-AMI initial admission. Multivariate Cox proportional hazard models, adjusted for baseline demographics and common comorbidities were used to compare risk of mortality at thirty-days and five-years from initial admission to hospital for all PMI patient compared with routine AMI admission, and then subgroup analyses compared outcomes according to classification of operations (orthopaedic, vascular, upper and lower gastrointestinal and urological).Results We identified 2,373 episodes of type-one PMI out of 472,278 eligible patients (0.5%). PMI was most frequently orthopaedic (53%) or vascular (15%), PMI patients were significantly older in median years compared with all-comers (IQR) (80 [73-87] vs. 70 [59-80], especially older in orthopaedic surgery (85 [79-90]). PMI patients were more frequently female compared with all-comers (47% vs. 33%) ( table 1). PMI was less frequently a STEMI compared with all-comers (18% vs. 42%). PMI patients received more conservative AMI care, with lower rates of P2Y12 inhibitors (76% vs. 90%), lower rates of invasive angiography (28% vs. 77%) and lower rates of revascularisation by either PCI or CABG (16% vs. 56%) (table 2). Adjusted risk of thirty-day all-cause mortality was significantly higher in PMI patients at thirty-days (1.60 [95% CIs 1.41-1.80]) and at five-years of follow-up (1.56 [1.46-1.67]) (both P<0.001) (figures 1 and 2). Discounting mortality prior to hospital discharge in a landmark analysis, PMI patients still had elevated all-cause mortality at five-years (1.36 [1.20-1.48), P<0.001).Conclusions Type-one perioperative MI is a more common than initially thought, especially frequent in orthopaedic and vascular surgery. It arises in a frailer cohort and is treated conservatively with low rates of secondary prevention and invasive investigation and treatments. All-cause mortality at thirty-days and five-years is significantly higher than all-cause MI patients, and this persists at five-years even when excluding patients that die prior to discharge.Perioperative MI comprises vascular, orthopaedic, upper & lower GI and urological proceduresAbstract 331 Figure 1Kaplan-Meier unadjusted survival curve for perioperative MI patients compared with AMI all-comersPerioperative MI comprises vascular, orthopaedic, upper gastrointestinal, lower gastrointestinal and urological proceduresAbstract 331 Figure 2Kaplan-Meier unadjusted survival curve for perioperative MI patients compared with AMI all-comers, stratified by operation typeAbstract 331 Table 1Baseline demographics of perioperative AMI patients, stratified by operation type, compared to all-comer AMI populationVariablesAll-comers (n=469,905)All perioperative MI (n=2,373)T & O (n=1,260)Vascular (n=360)Age, years, median (IQR)70 (59-80)80 (73-87)85 (79-90)75 (68-82)Ethnicity- White (%)321,708/354,767 (91)1,666/1,756 (95)878/907 (97)205/268 (95)Pulmonary oedema (%)15,319/311,542 (5)161/1,446 (11)75/754 (10)28/220 (12)ECG ST changes (%)392,093/461,419 (85)2,014/2,294 (88)1,062/1,220 (87)314/349 (90)Peripheral vascular disease (%)17,687/422,388 (4)242/2,230 (11)66/1,195 (6)124/339 (37)STEMI (%)195,147/469,905 (42)428/2,373 (18)159/1,260 (13)80/360 (22)Cardiac arrest (%)30,126/457,886 (7)300/2,324 (13)120/1,237 (10)48/347 (14Abdominal is a composite of upper gastrointestinal, lower gastrointestinal and urological proceduresAbstract 331 Table 2Management strategy and clinical outcome comparison between perioperative AMI patients compared to all-comer AMIVariablesAll-comers (n=469,905)All perioperative MI (n=2,373)T & O (n=1,260)Vascular (n=360)P2Y12 inhibitor (%)416,141/461,021 (90)1,774/2,338 (76)915/1,245 (73)282/352 (80)Inpatient coronary angiogram (%)351,215/456,851 (77)657/2,309 (28)226/1,224 (18)157/353 (44)Inpatient PCI (%)202,981/467,169 (54)355/2,365 (15)97/1,207 (8)91/358 (20)Inpatient mortality (%) (all-cause)23,871/469,905 (5)518/2,373 (22)280/1,260 (22)82/360 (23)30 day mortality (%) (all-cause)30,431/469,905 (6)499/2,373 (21)269/1,260 (21)84/360 (23)One year mortality (%) (all-cause)68,223/469,905 (15)916/2,373 (39)529/1,260 (42)128/360 (36)MACE (%) (inpatient)27,333/469,905 (6)538/2,373 (23)288/1,260 (23)85/360 (24)Abdominal is a composite of upper gastrointestinal, lower gastrointestinal and urological procedures