Spectrum From Acute Myocardial Injury to Infarction Among People With Human Immunodeficiency Virus Seeking Emergency Care in the United States: Presentations, Provider Responses, and Clinical Outcomes Authors: Rebecca A Abelman, Brian M Mugo, Claudia G Durbin, Sophia Campbell, Sayon Dutta, Dustin McEvoy, Emily S Lau, Sophia Zhao, Sara L Stockman, Sarah M Chu, Markella V Zanni Published: 2025 ABSTRACT AbstractBackgroundIn the United States (US), people with human immunodeficiency virus (PWH) have an increased risk of myocardial infarction, including type 1 myocardial infarction (T1MI) and type 2 myocardial infarction (T2MI). Presentations and clinical trajectories of PWH experiencing acute myocardial injury (AMI) have not been well characterized.MethodsLeveraging electronic health records (EHRs) from a US academic medical center, we identified PWH presenting to the emergency department from 2015 to 2019 with a troponin T ≥99th percentile. Presentations were adjudicated as AMI, T2MI, or T1MI. Clinical presentations, provider-level responses, and ensuing clinical outcomes (post–index event) were compared. Among PWH with AMI or T2MI, observed incidence of ensuing major adverse cardiovascular event (MACE) was evaluated using the cumulative incidence function by Aalen estimator. Adjusted cause-specific Cox proportional hazards models were used to assess the association between presentation type and ensuing MACE.ResultsAmong 79 cases analyzed, presentations of AMI and T2MI were more common than T1MI (29.1% and 64.6% vs 6.3%, respectively). Infection represented the most common event trigger for AMI and T2MI. Among PWH presenting with AMI versus T2MI, there was no difference in risk of ensuing MACE (adjusted hazard ratio, 1.14 [95% confidence interval, .48–2.71]). The proportion of cases of AMI versus T2MI not categorized with any cardiovascular disease–related diagnosis code differed significantly (91% vs 53%, P = .001).ConclusionsAmong US PWH presenting for emergency care, AMI was infrequently coded in the EHR. AMI and T2MI were associated with comparable rates of ensuing MACE. Enhanced recognition/documentation of AMI among PWH will facilitate development of preventive care approaches. People with human immunodeficiency virus 1 (HIV-1, or HIV) face a 2-fold increased risk of atherosclerotic cardiovascular disease (CVD), including myocardial infarction and stroke, versus people without known HIV [1]. In a retrospective study within the Center for AIDS Research Network of Integrated Clinical Systems (CNICS) cohort, investigators found that among people with HIV (PWH) in the United States (US) who have a myocardial infarction (dynamic elevation in troponin coupled with at least 1 symptom or sign of myocardial ischemia), 50.4% have a type 1 myocardial infarction (T1MI, caused by atherothrombosis) and 49.6% have a type 2 myocardial infarction (T2MI, caused by myocardial oxygen supply/demand mismatch in the absence of atherothrombosis) [2]. Meanwhile, studies in the general population suggest that patients with T2MI—and even patients with acute myocardial injury (AMI; dynamic elevation in troponin absent symptoms/signs of myocardial ischemia)—incur increased risk of ensuing adverse events, including death [3]. Whether AMI is similarly associated with adverse ensuing outcomes among PWH remains unknown. Notably, given heterogeneity in the etiopathology of AMI and T2MI, cardiology society guidelines offer no proscriptive strategy for treating patients with these presentations, postevent, to improve outcomes [4]. Taken together, these observations underscore the need for a better comparative understanding of myocardial injury and myocardial infarction presentations and outcomes among PWH. To address this knowledge gap, we conducted a retrospective analysis of PWH who presented to the emergency department (ED) of a large, Boston-area academic medical center (Mass General Brigham [MGB]) with a troponin T ≥99th percentile and were determined by detailed chart review to have either AMI, T2MI, or T1MI (defined as per Figure 1). Our primary aim was to compare participant characteristics and index event triggers, clinical care and documentation practices, and long-term ensuing clinical outcomes (post–index event) among PWH grouped by presentation type. Spectrum of acute heart muscle damage: definitions and theoretical etiopathology. Definitions and etiopathology adapted from Thygesen et al, Circulation 2018; 138:e618–51 [5], with deviations as described by double asterisk. **Participants without apparent 20% rise or fall in troponin who were noted to have both symptom and electrocardiographic or imaging sign of myocardial ischemia (with no evidence of atherothrombosis) were classified as having type 2 myocardial infarction (vs chronic myocardial injury) due to acuity of emergency department presentation. Abbreviations: c/w, consistent with; ECG, electrocardiogram; O2, oxygen; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction; Tn, troponin; ULN, upper limit of normal. METHODS Study Design and Analytic Cohort The study entailed retrospective chart review and was approved by the MGB Institutional Review Board. MGB electronic health records were screened for identification of adults (>18 years) with HIV who presented to 1 of the hospital system EDs between 31 May 2015 and 1 January 2019 with a resulted troponin T ≥99th percentile (positive troponin) ordered prior to ED departure. Manual chart review was performed to confirm HIV status and positive troponin. As high-sensitivity troponin had not yet been adopted in clinical practice during the studied time period, the 99th percentile range of the particular assay employed was used. If an individual with HIV presented multiple times to the ED with a positive troponin value, the first episode of positive troponin in the ED setting was selected (index event), even if it occurred outside of the 2015–2019 window. Clinical Data Extraction For each participant in the study sample, study staff assembled electronic data packets. Packets included admission and discharge notes for the index event, as well as other data related to the index event including laboratory values, electrocardiograms, and cardiovascular imaging and/or intervention reports. Teams of clinician-investigators used standardized data collection forms to record participants’ clinical characteristics, event triggers, and medical care delivered by providers. Additionally, for each participant, teams of clinician-investigators manually searched the electronic health records (EHRs) for evidence of ensuing prespecified clinical events (myocardial infarction, stroke, death) and noted timing of occurrence after the circumscribed index event, although these events were not adjudicated. Cause of death, if available, was also recorded. Clinician-investigators recorded any International Classification of Diseases (ICD) diagnosis codes relevant to cardiovascular health applied to participants’ cases. For participants with either T1MI or T2MI, we calculated the proportions of participants whose cases received (1) an ICD code referencing “myocardial infarction”; (2) a CVD-related ICD code not referencing “myocardial infarction” (ie, a more descriptive, less specific code); or (3) no CVD-related ICD code. For participants with AMI, no cases received an ICD code referencing AMI. Thus, we calculated the proportions of participants whose cases received any CVD-related ICD code or no such code. Adjudication Processes for Characterization of AMI, T2MI, and T1M1 Two physician-investigators (R. A. A. and M. V. Z.) independently adjudicated index events to characterize whether presentations were consistent with chronic myocardial injury (static troponin elevation absent symptoms/signs of myocardial ischemia), AMI, T2M1, or T1M1. The Fourth Universal Definition of Myocardial Infarction was used [5], with 1 minor modification engendered by patients’ acute presentation for emergency-level care (Figure 1). In cases of discrepant independent adjudication by primary reviewers, a physician-investigator board certified in cardiology (E. S. L.) performed the tie-breaking adjudication. Statistical Analysis Demographic, clinical, and laboratory characteristics were summarized using counts with percentages for categorical variables and medians with interquartile ranges (IQRs) for continuous variables. Data comparisons between groups of PWH with either AMI or T2MI were performed via Pearson χ2 test, Fisher exact test, or Wilcoxon test, as appropriate. The primary outcome was a subsequent major adverse cardiovascular event (MACE), which included subsequent myocardial infarction, stroke, or cardiac-related death. The secondary outcome was MACE and all-cause death. For the primary outcome, the cumulative incidence functions were estimated by the Aalen estimator for the probability of subdistribution of MACE, treating noncardiac death as a competing risk. Cause-specific hazard (ie, hazard ratio [HR]) for ensuing outcome among PWH presenting with AMI versus T2MI was estimated from Cox proportional hazards models. For the secondary outcome, the cumulative incidence was presented as 1 minus Kaplan-Meier survival estimates and the HR was estimated from Cox proportional models. No regression analysis was performed for the group of T1MI due to the small sample size. Models were adjusted for age (years) and sex at birth. Covariates were selected a priori given their known association with subsequent cardiac events. Schoenfeld residuals were examined, and the proportional hazards assumption was met for each model. Individuals were censored at loss to follow-up or at the end of the follow-up period if they did not have an event of interest by December 2023. RESULTS Adjudication Outcomes and Baseline Characteristics by Index Presentation Type Of the 149 participants’ presentations identified for screening, 25 were excluded for not having HIV, 8 were excluded for not having a sufficiently elevated troponin level, 1 was excluded for having a troponin drawn after leaving the ED, and 4 were excluded for incomplete data, leaving 111 participants’ presentations to be analyzed (Supplemental Figure 1). Among these, 32 cases were adjudicated as chronic myocardial injury and thus excluded from the analysis. The final analytic cohort was comprised of cases of 79 PWH: 23 (29.1%) with AMI, 51 (64.6%) with T2MI, and 5 (6.3%) with T1MI. Baseline characteristics of the 79 PWH overall and grouped by presentation type are presented in Table 1. Comparison of characteristics between PWH presenting with AMI versus T2MI revealed that hypertension prevalence was higher among PWH with T2MI (80.4% vs 56.5%, P = .03). A trend toward higher prevalence of dyslipidemia among PWH with T2MI was also noted (45.1% vs 26.1%, P = .12). With respect to HIV-associated factors, there was a trend toward lower CD4+ T-cell counts in those who presented with AMI versus T2MI (median [Q1, Q3], 217 [121, 670] cells/μL vs 466 [210, 774] cells/μL). No statistically significant between-group differences in the percentage of participants with undetectable HIV-1 RNA were observed. Table 1.Participant Characteristics at the Index EventCharacteristicFull GroupT1MIT2MIAMI P Value(T2MI vs AMI)No. of participants (%)795 (6.3)51 (64.6)23 (29.1)Age, y, mean ± SD59 ± 1163 ± 1159 ± 1158 ± 10.67Female sex at birth17 (21.5)1 (20.0)12 (23.5)4 (17.4).76Racea.72 White32 (43.2)2 (50.0)19 (38.8)11 (52.4) African American/Black31 (41.9)2 (50.0)22 (44.9)7 (33.3) Asian1 (1.4)0 (0.0)1 (2.0)0 (0.0) Other10 (13.5)0 (0.0)7 (14.3)3 (14.3)Ethnicityb1.00 Hispanic12 (17.4)1 (20)8 (17.4)3 (16.7)History of myocardial infarction19 (24.1)1 (20.0)15 (29.4)3 (13.0).15History of PCI, stent, or CABG14 (17.7)1 (20.0)11 (21.6)2 (8.7).32History of stroke15 (19.0)2 (40.0)6 (11.8)7 (30.4).09Cigarette smoking.17 Current21 (26.6)0 (0.0)16 (31.4)5 (21.7) Former30 (38.0)4 (80.0)20 (39.2)6 (26.1) Never28 (35.4)1 (20.0)15 (29.4)12 (52.2)Cocaine usec.47 Current12 (15.4)0 (0.0)10 (19.6)2 (9.1) Former22 (28.2)1 (20.0)15 (29.4)6 (27.3) Never44 (56.4)4 (80.0)26 (51.0)14 (63.6)Hypertension diagnosis58 (73.4)4 (80.0)41 (80.4)13 (56.5).03Diabetes mellitus diagnosis27 (34.2)3 (60.0)16 (31.4)8 (34.8).77Taking antihyperglycemic medication19 (24.1)3 (60.0)12 (23.5)4 (17.4).76Dyslipidemia diagnosis33 (41.8)4 (80.0)23 (45.1)6 (26.1).12Taking lipid-lowering medication39 (49.4)4 (80.0)26 (51.0)9 (39.1).34Mental health ailment diagnosis33 (41.8)2 (40.0)20 (39.2)11 (47.8).49Taking medication for a mental health ailment35 (44.3)1 (20.0)21 (41.2)13 (56.5).22ESRD on dialysis16 (20.3)1 (20.0)11 (21.6)4 (17.4).76Creatinine, mg/dL, median (Q1, Q3)1.90 (1.13, 5.01)3.27 (1.93, 6.05)1.78 (1.04, 5.64)1.89 (1.2, 3.18).66eGFR, mL/min/1.73 m2, median (Q1, Q3)38 (13, 74)16 (10, 47)38 (11, 84)43 (19, 66).62Hemoglobin, g/dL, mean ± SD11.3 ± 2.611.8 ± 2.111.2 ± 2.711.5 ± 2.7.64BMI, kg/m2, median (Q1, Q3)d24.7 (20.9, 28.2)24.0 (18.9, 29.0)25.0 (21.1, 27.8)24.4 (18.9, 29.3).72Glucose, mg/dL, median (Q1, Q3)112 (96, 157)200 (136, 450)112 (95, 155)108 (88, 140).82LDL-C, mg/dL, median (Q1, Q3)e74 (60, 99)153 (60, 242)73 (59, 93)78 (69, 103).28Current ART use72 (91.1)5 (100.0)45 (88.2)22 (95.7).42Total duration of ART use, y, median (Q1, Q3)f14 (6, 20)15 (2, 25)15 (8, 20)11 (1, 20).33Most recent HIV RNA undetectableg51 (66.2)3 (60.0)34 (68.0)14 (63.6).72Most recent CD4+ T-cell count, cells/μL, median (Q1, Q3)h399 (190, 675)417 (399, 550)466 (210, 774)217 (121, 670).13Chronic active hepatitis Bi4 (5.1)0 (0.0)3 (5.9)1 (4.4)1.00HCV positivity27 (34.2)0 (0.0)20 (39.2)7 (30.4).47Data are presented as No. (%) unless otherwise indicated.Abbreviations: AMI, acute myocardial injury; ART, antiretroviral therapy; BMI, body mass index; CABG, coronary artery bypass graft; eGFR, estimated glomerular filtration rate; ESRD, end-stage renal disease; HCV, hepatitis C virus; HIV, human immunodeficiency virus; LDL-C, low-density lipoprotein cholesterol; PCI, percutaneous intervention; Q, quartile; SD, standard deviation; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction.Data elements missing by category: a5 (1 T1MI, 2 T2MI, 2 AMI). b10 (5 T2MI, 5 AMI). c1 (1 AMI). d4 (1 T1MI, 1 T2MI, 2 AMI). e26 (1 T1MI, 13 T2MI, 12 AMI). f36 (2 T1MI, 22 T2MI, 12 AMI). g2 (1 T2MI, 1 AMI). h7 (6 T2MI, 1 AMI). i1 (1 T1MI). Participant Characteristics at the Index Event Data are presented as No. (%) unless otherwise indicated. Abbreviations: AMI, acute myocardial injury; ART, antiretroviral therapy; BMI, body mass index; CABG, coronary artery bypass graft; eGFR, estimated glomerular filtration rate; ESRD, end-stage renal disease; HCV, hepatitis C virus; HIV, human immunodeficiency virus; LDL-C, low-density lipoprotein cholesterol; PCI, percutaneous intervention; Q, quartile; SD, standard deviation; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction. Data elements missing by category: a5 (1 T1MI, 2 T2MI, 2 AMI). b10 (5 T2MI, 5 AMI). c1 (1 AMI). d4 (1 T1MI, 1 T2MI, 2 AMI). e26 (1 T1MI, 13 T2MI, 12 AMI). f36 (2 T1MI, 22 T2MI, 12 AMI). g2 (1 T2MI, 1 AMI). h7 (6 T2MI, 1 AMI). i1 (1 T1MI). Triggers for Index Event Infection was the most common trigger for both AMI (39.1%) and T2MI (47.1%) (Table 2). For T1MI, triggers were divided between infection, hypertension, and severe metabolic derangement. Hypotension was more frequently a trigger for AMI versus T2MI (P = .01), while heart failure decompensation and respiratory decompensation were more frequently triggers for T2MI versus AMI (P = .01 and P = .049, respectively) (Table 2). Table 2.Possible Clinical Triggers for Index EventPossible TriggerFull GroupT1MIT2MIAMI P Value(AMI vs T2MI)No. of participants795 (6.3)51 (64.6)23 (29.1)Pulmonary embolism2 (2.5)0 (0.0)2 (3.9)0 (0.0)1.00Infection34 (43.0)1 (20.0)24 (47.1)9 (39.1).53Among those with infection as triggera: Bacteremia/sepsis14 (41.2)0 (0.0)9 (37.5)5 (55.6).44 Pulmonary/pneumonia18 (52.9)0 (0.0)17 (70.8)1 (11.1).004 UTI6 (17.7)1 (100.0)3 (12.5)2 (22.2).60 Other10 (29.4)0 (0.0)5 (20.8)5 (55.6).09Substance use9 (11.4)0 (0.0)5 (9.8)4 (17.4).45Tachycardia14 (17.7)0 (0.0)9 (17.7)5 (21.7).75Bradycardia1 (1.3)0 (0.0)1 (2.0)0 (0.0)1.00Heart failure decompensation12 (15.2)0 (0.0)12 (23.5)0 (0.0).01Hypertension17 (21.5)2 (40.0)8 (15.7)7 (30.4).21Hypotension13 (16.5)0 (0.0)5 (9.8)8 (34.8).01Respiratory decompensation21 (26.6)0 (0.0)18 (35.3)3 (13.0).049Gastrointestinal bleed4 (5.1)0 (0.0)1 (2.0)3 (13.0).09Severe metabolic derangement24 (30.4)2 (40.0)16 (31.4)6 (26.1).65Data are presented as No. (%) unless otherwise indicated.Abbreviations: AMI, acute myocardial injury; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction; UTI, urinary tract infection. aMore than 1 trigger possible for each index event. Possible Clinical Triggers for Index Event Data are presented as No. (%) unless otherwise indicated. Abbreviations: AMI, acute myocardial injury; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction; UTI, urinary tract infection. aMore than 1 trigger possible for each index event. Provider Clinical Response The majority of studied PWH (across all presentations) (97.5%) were admitted to the hospital for further care following the index ED presentation. There were key differences in clinical care rendered by providers for PWH across groups: Lower proportions of PWH with AMI versus T2MI received inpatient cardiology consultation (P = .002), received cardioprotective medications on discharge (P = .002), or were seen by an outpatient cardiologist within 1 year of discharge (P = .003) (Table 3). Table 3.Clinical Care Received Following Emergency Department EvaluationTreatmentFull GroupT1MIT2MIAMI P Value(T2MI vs AMI)No. of participants795 (6.3)51 (64.6)23 (29.1)Admitted to the inpatient hospital wards from the ED77 (97.5)5 (100.0)50 (98.0)22 (95.7).53On aspirin, beta blocker, and statin on admissiona20 (26.0)1 (20.0)17 (34.0)2 (9.1).03On aspirin, beta blocker, and statin on dischargeb29 (39.7)4 (80.0)23 (48.9)2 (9.5).002Received inpatient cardiology consultationc26 (36.6)5 (100.0)20 (43.5)1 (5.0).002Seen by an outpatient cardiologist within 1 y of dischargeb31 (42.5)3 (60.0)25 (53.2)3 (14.3).003Data are presented as No. (%) unless otherwise indicated.Abbreviations: AMI, acute myocardial injury; ED, emergency department; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction.Data elements missing by category: a2 (1 T2MI, 1 AMI) not admitted to the hospital. b6 (4 T2MI, 2 AMI) died in hospital. c8 (5 T2MI, including 1 not admitted to the hospital and 4 who died during hospitalization; 3 AMI, including 1 not admitted in the hospital and 2 who died during hospitalization). Clinical Care Received Following Emergency Department Evaluation Data are presented as No. (%) unless otherwise indicated. Abbreviations: AMI, acute myocardial injury; ED, emergency department; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction. Data elements missing by category: a2 (1 T2MI, 1 AMI) not admitted to the hospital. b6 (4 T2MI, 2 AMI) died in hospital. c8 (5 T2MI, including 1 not admitted to the hospital and 4 who died during hospitalization; 3 AMI, including 1 not admitted in the hospital and 2 who died during hospitalization). Time to Ensuing MACE and Composite Outcome (Distinct From Index Event Presentation) Overall, there was no statistically significant differences in risk of ensuing MACE or MACE plus all-cause death among PWH who presented with AMI versus T2MI (Figure 2). In the unadjusted Cox proportional hazards models evaluating time to MACE, with T2MI as the reference, AMI had similar hazards for an ensuing MACE event (HR, 1.09 [95% confidence interval {CI}, .46, 2.59]). In the adjusted analysis, results were comparable, with AMI having a similar risk of MACE compared to T2MI (HR, 1.14, [95% CI, .48, 2.71]). Probability of ensuing major adverse cardiovascular event after index presentation of acute myocardial injury and type 2 myocardial infarction. Abbreviations: CI, confidence interval; MACE, major adverse cardiovascular event; MI, myocardial infarction. For the composite outcome of MACE or death (any cause), the unadjusted and adjusted Cox proportional hazards models yielded similar results as models for time to MACE. AMI, as compared to T2MI, had an unadjusted HR of 1.01 (95% CI, .51, 2.01) and an adjusted HR of 1.05 (95% CI, .53, 2.10) for ensuing MACE plus all-cause mortality. Coding of Events in the EHRs There were key differences with respect to how index events were documented in the EHRs via ICD codes (Supplemental Figure 2). Among PWH with T1MI, 80% of cases were coded as “myocardial infarction” and 20% of cases received a CVD-related ICD code not referencing “myocardial infarction” (ie, a more descriptive code, such as “elevated troponin”). Among PWH with T2MI, 14% of cases were coded as “myocardial infarction,” 33% of cases received a CVD-related ICD code not referencing “myocardial infarction,” and 53% of cases received no CVD-related ICD code. Finally, among PWH with AMI, no cases were coded as “myocardial injury,” 9% of cases received a CVD-related ICD code not referencing “myocardial injury,” and 91% of cases received no CVD-related ICD code. There was a statistically significant difference in the proportion of cases of AMI versus T2MI not categorized with any CVD-related ICD code (P = .001) (Supplemental Figure 2). DISCUSSION Evaluating PWH who presented for emergency care and were found to have acute myocardial injury or infarction, we found that presentations of AMI and T2MI were more common than presentations of T1MI. While participants’ baseline data did not allow us to distinguish a clear set of characteristics for each event type, select traditional CVD risk factors were more prevalent among those presenting with T2MI versus AMI. Conversely, CD4+ T-cell counts tended to be lower among PWH with AMI versus T2MI. Infection represented the most common trigger for both AMI and T2MI. In terms of postdischarge follow-up, outpatient cardiology follow-up visit within 12 months infrequently occurred for PWH with AMI. However, among PWH presenting with AMI versus T2MI, there was no significant difference in risk of ensuing MACE or in risk of MACE plus all-cause mortality. Finally, AMI and T2MI were infrequently coded correctly in EHR summaries of the index event hospital stay, with potential consequences for future secondary cardiovascular disease prevention efforts. While researchers analyzing CNICS data have usefully compared baseline characteristics and event triggers among PWH experiencing either T2MI or T1MI, no prior studies have compared PWH presenting for emergency-level care and found to have AMI or T2MI. In our cohort, PWH with T2MI had a higher prevalence of select traditional CVD risk factors (eg, hypertension) while PWH with AMI tended to have lower CD4+ T-cell counts. Meanwhile, infection represented the most common trigger for both presentations. For comparison, in CNICS studies, there was noted to be an increased prevalence of traditional CVD risk factors (eg, dyslipidemia) among patients with T1MI and an increased prevalence of HIV-specific risk factors (eg, lower CD4+ T-cell count) among patients with T2MI [2]. Juxtaposing our MGB findings with CNICS findings, we appreciate that PWH with a greater burden of traditional CVD risk factors may be predisposed to experience focal infarction in the setting of a “demand” event (such as infection) whereas PWH with predominantly HIV-specific risk factors (such as immune dysfunction) may experience injury absent infarction. In our study, most PWH presenting for emergency-level care did not receive inpatient cardiology consultation, cardiovascular disease prevention medications, or timely outpatient cardiology follow-up for AMI or T2MI. Among PWH presenting with AMI or T2MI, underlying etiopathologic contributors may or may not include subclinical atherosclerosis in the epicardial coronary arteries, coronary artery vasospastic diathesis, or coronary microvascular dysfunction (impaired ability to augment oxygenated blood flow through the coronary microvascular bed in response to enhanced oxygen demand) (Figure 1). For example, a patient with either AMI or T2MI may have experienced myocardial damage triggered by a “demand” event (eg, infection) in isolation, or by a “demand” event superimposed on underlying vascular pathology. In this context, for patients with these presentations, no unified clinical approach has been espoused in medical society guidelines. It may be reasonable not to reflexively offer such patients inpatient cardiology consultation (resource constraints) or cardiovascular disease prevention medications (especially if lack of clear indication). Future implementation science studies may usefully assess whether deployment of CVD prevention strategies among PWH who present with AMI or T2MI improve clinical outcomes. Our work is the first to show that among PWH with AMI, risk of ensuing longer-term adverse outcomes, including MACE or MACE plus all-cause mortality, are on par with those among PWH with T2MI. Prior studies comparing patients with AMI versus T2MI or comparing patients with different forms of myocardial infarction have focused exclusively on an all-cause mortality outcome. In a study of patients without known HIV discharged from a US academic medical center with a diagnosis of either myocardial injury or T2MI, subsequent mortality rates were 7.4% versus 4.4%, respectively (P = .14) [6]. Meanwhile, data from CNICS revealed that among PWH, mortality rates were significantly higher after T2MI as compared to T1MI (1-, 3-, and 5-year mortality: 39%, 52%, and 62% vs 15%, 22%, and 30%, respectively) [7]. The relationship between either AMI or T2MI with all-cause mortality may be partially explained by the presence of comorbid conditions predisposing to catastrophic “demand ischemia” events, such as sepsis. However, our analysis reveals a comparable, relatively high rate of ensuing MACE among PWH presenting with either AMI or T2MI. Analogously, a recent Randomized Trial to Prevent Vascular Events in HIV (REPRIEVE) Trial analysis showed that among a relatively healthy CVD prevention cohort of PWH (without end-stage renal disease or active cancer), chronic myocardial injury portended an increased risk of ensuing MACE [8]. Taken together, these findings suggest that among PWH, chronic or acute myocardial injury and T2MI presentations may reflect underlying cardiovascular pathology, which, if untreated, predispose to MACE. Among PWH who have experienced chronic or acute myocardial injury or T2MI, research is needed to assess whether characterizing cardiovascular pathology and subsequently addressing modifiable CVD risk factors may mitigate the risk of ensuing cardiovascular events. With respect to documentation, among PWH seeking emergency care, AMI and T2MI were infrequently coded correctly in EHR summaries of the index event. A prior study examining EHRs of patients without known HIV illustrated that electronic coding of even acute myocardial infarction is highly insensitive as compared to analysis of chart-abstracted data [9]. Separate studies applying chart-based adjudication strategies have identified frequent miscoding of myocardial infarction presentations [6]. Accurate coding of chronic myocardial injury, AMI, T2M1, and T1MI likely represents an important first step in facilitating optimal clinical care of patients and reproducible, high-impact EHR-based research. From a clinical standpoint, if outpatient clinicians note (based on coding) that an individual with HIV experienced an inpatient myocardial injury or infarction event, such clinicians may be more likely to consider characterizing cardiovascular pathology and addressing modifiable CVD risk factors. From a research standpoint, accurate coding will facilitate observational assessment of whether real-world clinical strategies may protect against future CVD events, particularly among PWH with AMI or T2MI. Limitations of our study are as follows: First, given the relatively small sample size and the fact that our study focused on PWH from a single, urban US region, findings may not be widely generalizable. Second, our study did not include PWH presenting for emergency care found to have elevations in high-sensitivity troponin. In this regard, little is known about real-world outcomes of PWH with very mild AMI, such as can be detected with the newer assays. Third, our real-world study classified presentations based on patterns emerging from troponin values drawn by ED clinicians. However, independent decision-making by clinicians may have introduced variability. That is, if a clinician checked a troponin once in a patient without new-onset cardiac symptoms or signs, that patient would be deemed to have chronic myocardial injury and would not be included in the present analysis. However, if the clinician had opted to check serial troponins in this same patient, and if serial values reflected a ≥20% change in troponin, that patient would be classified as having experienced AMI. Fourth, our study featured a small number of PWH presenting for emergency care and found to have T1MI events, limiting our comparisons of outcomes among PWH with AMI and T2MI versus T1MI. The relatively low proportion of T1M1 events we observed may be due, in part, to patients with T1MI being diagnosed in the field or in clinic and next routed, per protocol [10], directly to the care of the cardiac catheterization team, circumventing ED evaluation. Fifth, clinical decision-making may have been in response to past medical history as well as ED presentation. Additionally, with respect to assessment of cardiology follow-up visits completed, it is possible that either within-system referrals were placed and appointments were not kept or that appointments taking place at sites with health record systems inaccessible to chart reviewers were missed. Finally, ensuing events were not formally adjudicated but were based off chart review, potentially causing select events to be missed, especially if these occurred at an outside facility. Overall, our study findings build on previously published work about myocardial infarction risk in PWH [2, 7, 11, 12], and, via a unique study design, highlights novel clinical observations: among PWH who presented to the ED of a large, urban academic medical center with a positive troponin, AMI and T2MI were highly prevalent, and there were no significant differences in risk of ensuing MACE or in risk of MACE plus all-cause mortality among PWH with AMI versus T2MI. Nevertheless, low-level attention was paid to event documentation and implementation of cardiovascular disease prevention care for PWH with AMI and T2MI. This observation may reflect providers’ perceptions that myocardial injury and T2MI are less clinically relevant than events that may have precipitated ED presentation (eg, severe infections). Nevertheless, our findings suggest potential utility in apprising clinical providers that myocardial injury and T2MI may be harbingers of ensuing MACE. Coding these presenting events and studying mechanisms underlying event-to-subsequent-MACE relationships represents an important step toward determining whether targeted CVD prevention strategies may benefit PWH who present with either AMI or T2MI. Supplementary Material