Abstract
Introduction:
This study evaluated surgical outcomes of infective endocarditis (IE), with particular attention to the impact of intravenous drug use (IVDU).
Methods:
Adult patients undergoing surgery for IE between 2011 and 2018 at a single center were included and stratified by IVDU. The primary outcome was overall survival. Secondary outcomes included postoperative complications and hospital readmissions. Kaplan-Meier and multivariable Cox regression were utilized for unadjusted and risk-adjusted survival analyses, respectively. Cumulative incidence function curves were compared for hospital readmissions.
Results:
A total of 831 patients (mean age 55 years, 34.4% female) were operated on for IE, including 318 (38.3%) with IVDU. Cultures were most commonly positive for streptococcus (25.2%), methicillin-sensitive Staphylococcus aureus (17.7%), enterococcus (14.3%), or methicillin-resistant Staphylococcus aureus (8.4%). The most common procedures included isolated aortic valve repair/replacement (18.8%), aortic root replacement (15.9%), mitral valve repair/replacement (26.7%), aortic and mitral valve replacement (8.4%), and tricuspid valve repair/replacement (7.6%). Mean follow-up was 3.4 ± 2.4 years. Overall 5-year survival was 64% and was similar between IVDU and non-IVDU. Multivariable analysis demonstrated that IVDU was not associated with mortality risk. IVDU patients displayed higher rates of all-cause readmission (61.6% vs 53.9%; P = .03), drug-use readmission (15.4% vs 1.4%; P < .001), and recurrent endocarditis readmission (33.0% vs 13.0%; P < .001).
Conclusions:
The majority of patients undergoing surgical treatment of IE are alive at 5-years although readmission rates are high. IVDU is not a risk factor for longitudinal mortality although patients with IVDU are at higher overall readmission risk, driven largely by greater readmissions for drug-use and recurrent endocarditis.
Keywords: cardiovascular pathology, surgical history, valve repair/replacement
1 |. INTRODUCTION
The profile of patients with infective endocarditis (IE) has shifted over the past decade though overall rates of IE have remained relatively constant.1–3 In particular, rates of IE attributable to intravenous drug use (IVDU) increased more than 12-fold from 2007 to 2017 and IVDU-associated IE has been estimated to account for approximately 42% of valve surgeries for IE.3 Prior work has examined the relationship between IVDU and IE surgical outcomes though results have been conflicting. Two recent meta-analyses explored IVDU and postoperative IE outcomes, finding equivalent 30-day mortality but inferior long-term survival with IVDU.4,5 Others, however, have found comparable 5- and 10-year postoperative survival between IVDU and non-IVDU.6 Given the public health burden of IE and the growing opioid epidemic as well as high rates of IVDU recidivism, accruing additional data on the outcomes in this challenging population is prudent.7,8 The aim of this study was to compare outcomes among patients undergoing surgery for IE, with a specific focus on long-term survival and readmissions in patients with and without a history of IVDU.
2 |. METHODS
2.1 |. Study population and outcomes
This study utilized institutionally-derived data from the Society of Thoracic Surgeons (STS) adult cardiac database supplemented with longitudinal data extracted from our electronic health record (EHR). All consecutive patients (≥18 years) undergoing surgery for IE between 2011 and 2018 at our center were included. This includes redo cases. Patients were stratified based on IVDU. The primary outcome was overall survival. Secondary outcomes included postoperative complications and hospital readmission. Length of stay represented postsurgical stay, including any inpatient stay for intravenous antibiotic administration. Readmissions were categorized as all-cause, cardiac, heart failure, drug-related, or endocarditis-related hospitalizations which were obtained by review of the EHR. For all outcomes, patients were censored at the time of their last follow-up visit as recorded in our EHR to account for potential loss to follow-up. This study was approved by the Institutional Review Board of the University of Pittsburgh.
2.2 |. Statistical analysis
Demographic and clinical characteristics are presented as frequency (percentage) for categorical variables and mean ± standard deviation for Gaussian continuous variables or median (interquartile range [IQR]) for non-Gaussian continuous variables. Normality was checked with the Kolmogorov-Smirnov test. The χ2 test or Fisher’s exact test was utilized for categorical variables and continuous variables were analyzed with t-tests if normally distributed and the Mann-Whitney U test if non-Gaussian. Multivariable Cox regression was utilized for risk-adjusted analysis for mortality. The multivariable model was constructed incorporating both univariate predictors as well as variables based on clinical relevance. All baseline characteristics were evaluated for potential inclusion in the models. Kaplan-Meier estimates and the log-rank test were used to compare longitudinal survival. A cumulative incidence of readmission model, accounting for the possibility of multiple readmissions and utilizing death as a competing risk, was constructed for patients with and without IVDU. Statistical analyses were performed with version 9.4 SAS software (SAS Institute, Cary, NC).
3 |. RESULTS
3.1 |. Patient population
Between 2011 and 2018, 831 patients underwent surgery for IE with 318 (38.3%) having a history of IVDU and 508 (61.1%) with no history of IVDU (Table 1). Of note, 5 patients had questionable histories of drug use and were not included in the IVDU or no IVDU cohorts. The overall median age was 55 years with 286 (34.4%) being female; 296 (35.6%) had a prior valve surgery. Those with IVDU were significantly younger (36 vs 52 years; P < .0001) and had fewer comorbidities including diabetes (14.8% vs 36.0%; P < .0001) and chronic lung disease (17.0% vs 25.8%; P = .003). Those with IVDU less commonly had a prior valve surgery (28.3% vs 40.0%; P = .001). There were no significant differences in ejection fraction or cardiogenic shock between the groups (both P > .05). Those with IVDU more commonly underwent tricuspid valve repair/replacement while those without IVDU displayed higher rates of isolated aortic valve repair/replacement or mitral valve repair/replacement (P < .001). Culture results most commonly demonstrated streptococcus (25.2%); IVDU patients more commonly had polymicrobial (4.4% vs 1.0%), methicillin-resistant Staphylococcus aureus (MRSA) (12.6% vs 5.9%), methicillin-sensitive Staphylococcus aureus (MSSA) (25.8% vs 12.8%), and Gram negative (5.0% vs 2.0%; P < .001) species.
TABLE 1.
Baseline characteristics of patients treated for endocarditis in the overall population and in patients without and with a history of intravenous drug use (IVDU)
| Overall N = 831 |
No IVDU N = 508 |
IVDU N = 318 |
P value | |
|---|---|---|---|---|
| Age, y | 55 (37–66) | 62 (52–71) | 36 (29–50) | <.0001 |
| Female sex | 286 (34.4%) | 174 (34.3%) | 111 (34.9%) | .85 |
| Race | .002 | |||
| White | 745 (89.7%) | 471 (92.7%) | 271 (85.2%) | |
| Black | 71 (8.5%) | 30 (5.9%) | 40 (12.6%) | |
| Other | 15 (1.8%) | 7 (1.4%) | 7 (2.2%) | |
| Median income by zip code (US dollars) | 46505 (40097–52155) | 47229 (47362–55117) | 43838 (38257–49473) | <.001 |
| BMI, kg/m2 | 26.6 (23.1–32) | 27.5 (24–33.7) | 25.1 (22.2–29.2) | <.0001 |
| Dyslipidemia | 334 (40.2%) | 272 (53.5%) | 60 (18.9%) | <.0001 |
| Diabetes Mellitus | 230 (27.7%) | 183 (36.0%) | 47 (14.8%) | <.0001 |
| Dialysis | 95 (11.4%) | 59 (11.6%) | 35 (11.0%) | .79 |
| Hypertension | 482 (58.0%) | 347 (68.3%) | 132 (41.5%) | <.0001 |
| COPD | 188 (22.6%) | 131 (25.8%) | 54 (17.0%) | .003 |
| Serum albumin | 2.9 (2.4–3.3) | 2.9 (2.5–3.5) | 2.7 (2.2–3.2) | <.0001 |
| Serum creatinine, mg/dL | 1 (0.8–1.5) | 1.1 (0.8–1.6) | 0.9 (0.7–1.4) | <.0001 |
| Immunosuppression | 83 (10.0%) | 63 (12.4%) | 19 (6.0%) | .003 |
| Peripheral arterial disease | 122 (14.7%) | 92 (18.1%) | 30 (9.4%) | .001 |
| Cerebrovascular disease | 290 (34.9%) | 186 (36.6%) | 101 (31.8%) | .15 |
| Previous PCI | 41 (4.9%) | 33 (6.5%) | 8 (2.5%) | .01 |
| Previous CABG | 67 (8.1%) | 60 (11.8%) | 7 (2.2%) | <.0001 |
| Previous valve surgery | 296 (35.6%) | 203 (40.0%) | 90 (28.3%) | .001 |
| Family history of CAD | 89 (10.7%) | 58 (11.4%) | 31 (9.8%) | .45 |
| Previous MI | 150 (18.1%) | 105 (20.7%) | 44 (13.8%) | .01 |
| Cardiac presentation | .03 | |||
| No symptoms | 368 (44.3%) | 219 (43.1%) | 149 (46.9%) | |
| Symptoms unlike ischemia | 148 (17.8%) | 112 (22.1%) | 32 (10.1%) | |
| Stable angina | 6 (0.7%) | 4 (0.8%) | 2 (0.6%) | |
| Unstable angina | 26 (3.1%) | 18 (3.5%) | 8 (2.5%) | |
| NSTEMI | 42 (5.1%) | 25 (4.9%) | 17 (5.4%) | |
| STEMI | 7 (0.8%) | 2 (0.4%) | 5 (1.6%) | |
| Other | 234 (28.2%) | 128 (25.2%) | 105 (33.0%) | |
| Congestive heart failure | 368 (44.3%) | 230 (45.3%) | 135 (42.5%) | .43 |
| NYHA class | .51 | |||
| I or no heart failure | 471 (56.7%) | 282 (55.5%) | 187 (58.8%) | |
| II | 36 (4.3%) | 25 (4.9%) | 11 (3.5%) | |
| III | 130 (15.6%) | 84 (16.5%) | 44 (13.8%) | |
| IV | 194 (23.4%) | 117 (23.0%) | 76 (23.9%) | |
| Cardiogenic shock | 45 (5.4%) | 29 (5.7%) | 15 (4.7%) | .54 |
| Arrhythmia | 161 (19.4%) | 131 (25.8%) | 30 (9.4%) | <.0001 |
| Left ventricular EF (%) | 58 (53–60) | 58 (10–76) | 58 (53–60) | .82 |
| Status | .002 | |||
| Elective | 116 (14.0%) | 87 (17.1%) | 28 (8.8%) | |
| Urgent | 636 (76.5%) | 370 (72.8%) | 262 (82.4%) | |
| Emergent | 75 (9.0%) | 50 (9.8%) | 25 (7.9%) | |
| Emergent salvage | 4 (0.5%) | 1 (0.2%) | 3 (0.9%) | |
| STS PROM% | 3.73 (1.44–9.13) | 5.6 (2.25–11.04) | 2 (1.07–5.12) | <.001 |
| Periannular abscess | 76 (9.2%) | 52 (10.2%) | 24 (7.6%) | .193 |
| Concomitant CABG | 95 (11.4%) | 69 (13.6%) | 20 (6.3%) | .001 |
| Operation | <.001 | |||
| AVR | 156 (18.8%) | 104 (20.5%) | 52 (16.4%) | |
| MVR | 222 (26.7%) | 148 (29.1%) | 71 (22.3%) | |
| TVR | 63 (7.6%) | 11 (2.2%) | 52 (16.4%) | |
| PVR | 6 (0.7%) | 2 (0.4%) | 4 (1.3%) | |
| Aortic root | 132 (15.9%) | 96 (18.9%) | 35 (11.0%) | |
| AVR + MVR | 70 (8.4%) | 52 (10.2%) | 18 (5.7%) | |
| AVR + TVR | 15 (1.8%) | 5 (1.0%) | 10 (3.1%) | |
| MVR+TVR | 43 (5.2%) | 22 (4.3%) | 21 (6.6%) | |
| Aortic root+ MVR | 42 (5.1%) | 28 (5.5%) | 14 (4.4%) | |
| Aortic root +TVR | 10 (1.2%) | 5 (1.0%) | 5 (1.6%) | |
| AVR or TVR + PVR | 4 (0.5%) | 2 (0.4%) | 2 (0.6%) | |
| Triple valve | 23 (2.8%) | 11 (2.2%) | 12 (3.8%) | |
| Other | 45 (5.4%) | 22 (4.3%) | 22 (6.9%) | |
| Endocarditis culture | <.001 | |||
| Negative | 77 (9.3%) | 59 (11.6%) | 17 (5.4%) | |
| Streptococcus species | 209 (25.2%) | 141 (27.8%) | 66 (20.8%) | |
| Coagulase negative staph | 63 (7.6%) | 48 (9.5%) | 15 (4.7%) | |
| Enterococcus species | 119 (14.3%) | 84 (16.5%) | 35 (11.0%) | |
| Fungal | 20 (2.4%) | 7 (1.4%) | 13 (4.1%) | |
| Other | 25 (3.0%) | 20 (3.9%) | 5 (1.6%) | |
| Gram negative species | 26 (3.1%) | 10 (2.0%) | 16 (5.0%) | |
| Polymicrobial | 20 (2.4%) | 5 (1.0%) | 14 (4.4%) | |
| MRSA | 70 (8.4%) | 30 (5.9%) | 40 (12.6%) | |
| MSSA | 147 (17.7%) | 65 (12.8%) | 82 (25.8%) | |
| Unknown | 55 (6.6%) | 39 (7.7%) | 15 (4.7%) |
Abbreviations: AVR, aortic valve repair or replacement; BMI, body mass index; CABG, coronary artery bypass grafting; CAD, coronary artery disease; EF, ejection fraction; MI, myocardial infarction; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-sensitive Staphylococcus aureus; MVR, mitral valve repair or replacement; NSTEMI, non-ST elevation myocardial infarction; PCI, percutaneous coronary intervention; PROM, predicted risk of operative mortality; PVR, pulmonic valve replacement; STEMI, ST-elevation myocardial infarction; STS, Society of Thoracic Surgeons; TVR, tricuspid valve repair or replacement.
3.2 |. Outcomes
Patients with IVDU had a significantly longer postoperative length of stay (18 vs 15 days; P < .001) and were more likely to have new-onset atrial fibrillation (38.5% vs 31.1%; P < .001) (Table 2). Operative mortality was lower in the IVDU group (6.6% vs 10.4%) though this was not statistically significant (P = .06). Additionally, renal failure (9.1% vs 12.8%; P = .11) and reoperation (17.6% vs 18.1%; P = .86) were not statistically different between the IVDU and non-IVDU groups, respectively. Outcomes for patients who underwent mitral valve repair versus replacement are also shown (Table S1). The median follow-up time was 3.06 years (IQR: 1.4–5.0). One-year survival was 81.1% in non-IVDU and 81.5% in patients with IVDU (P = .77); 5-year survival rates were also similar at 67.6% in non-IVDU and 57.3% in IVDU (P = .22) (Figure 1). The cumulative incidence of all-cause readmission was similar at 1 year (45.9% IVDU vs 38.0% non-IVDU; P = .08) but diverged at 5 years with IVDU patients exhibiting greater rates of readmission (68.0% vs 56.3%; P = .01) (Figure 2). Although the overall absolute percentage of patients being readmitted was higher in the IVDU group, there were no differences in the number of readmissions or time to first readmission in IVDU versus non-IVDU groups (Table 3). Both drug-related readmission (15.4% vs 1.4%; P < .001) and endocarditis-related readmission (33.0% vs 13.0%) were higher among IVDU patients whereas heart failure readmission (24.6% vs 17.3%, P = 0.01) was more frequent in the non-IVDU group. We also examined follow-up echocardiogram data for patients undergoing mitral (Table S2) or aortic (Table S3) surgery.
TABLE 2.
Postoperative outcomes among patients undergoing surgery for infective endocarditis, with and without a history of IVDU
| Overall N = 831 |
No IVDU N = 508 |
IVDU N = 318 |
P value | |
|---|---|---|---|---|
| Median length of stay, d | 16 (10–25) | 15 (9–22) | 18 (11–28) | <.001 |
| Operative mortality | 75 (9.0%) | 53 (10.4%) | 21 (6.6%) | .06 |
| Blood product transfusion | 531 (63.9%) | 336 (66.1%) | 192 (60.4%) | .09 |
| Prolonged ventilation >24 h | 209 (25.2%) | 138 (27.2%) | 69 (21.7%) | .08 |
| Renal failure | 94 (11.3%) | 65 (12.8%) | 29 (9.1%) | .11 |
| Deep sternal wound infection | 2 (0.2%) | 2 (0.4%) | 0 (0.0%) | .53 |
| Sepsis | 27 (3.3%) | 19 (3.7%) | 8 (2.5%) | .34 |
| Pneumonia | 51 (6.1%) | 41 (8.1%) | 10 (3.1%) | .004 |
| Permanent stroke | 23 (2.8%) | 15 (3.0%) | 8 (2.5%) | .71 |
| Reoperation | 150 (18.1%) | 92 (18.1%) | 56 (17.6%) | .86 |
| New-onset atrial fibrillation | 213 (25.6%) | 158 (31.1%) | 54 (38.5%) | <.001 |
Abbreviation: IVDU, intravenous drug use.
FIGURE 1.
Kaplan-Meier analysis of overall survival in those with and without a history of intravenous drug use (IVDU)
FIGURE 2.
Cumulative incidence of all-cause readmission in patients with and without a history of IVDU. IVDU, intravenous drug use
TABLE 3.
Outcomes for patients undergoing surgery for infective endocarditis, with and without a history of IVDU; for each readmission type, the total number of readmissions for all patients, the mean number of readmissions per patient (with standard deviation), and the median time to the first readmission for each patient (with interquartile range, in years) are shown
| Overall N = 831 |
No IVDU N = 508 |
IVDU N = 318 |
P value | |
|---|---|---|---|---|
| Death | 268 (32.3%) | 158 (31.1%) | 108 (34.0%) | .39 |
| All-cause readmission | ||||
| Readmissions no. (%) | 472 (56.8%) | 274 (53.9%) | 196 (61.6%) | .03 |
| Mean number | 1.83 (3.13) | 1.79 (3.05) | 1.92 (3.26) | .09 |
| Median time to first | 0.88 (0.05–2.91) | 0.98 (0.05–3.17) | 0.73 (0.08–2.46) | .56 |
| Cardiac | ||||
| Readmissions no. (%) | 419 (50.4%) | 243 (47.8%) | 174 (54.7%) | .054 |
| Mean number | 1.41 (2.66) | 1.36 (2.50) | 1.51 (2.91) | .09 |
| Median time to first | 1.15 (0.08–3.27) | 1.22 (0.06–3.54) | 1.06 (0.11–2.76) | .47 |
| Heart failure | ||||
| Readmissions no. (%) | 181 (21.8%) | 125 (24.6%) | 55 (17.3%) | .01 |
| Mean number | 0.46 (1.45) | 0.53 (1.56) | 0.34 (1.27) | .01 |
| Median time to first | 2.29 (0.47–4.28) | 2.28 (0.28–4.79) | 2.33 (0.77–3.82) | .55 |
| Drug-related | ||||
| Readmissions no. (%) | 56 (6.7%) | 7 (1.4%) | 49 (15.4%) | <.001 |
| Mean number | 0.10 (0.44) | 0.03 (0.28) | 0.22 (0.60) | <.001 |
| Median time to first | 2.87 (1.17–4.94) | 3.27 (1.37–5.37) | 2.31 (0.90–3.99) | <.001 |
| Endocarditis-related | ||||
| Readmissions no. (%) | 171 (20.6%) | 66 (13.0%) | 105 (33.0%) | <.001 |
| Mean number | 0.34 (0.82) | 0.17 (0.51) | 0.61 (1.11) | <.001 |
| Median time to first | 2.21 (0.44–4.38) | 2.85 (0.64–4.98) | 1.72 (0.37–3.44) | <.001 |
Abbreviation: IVDU, intravenous drug use.
Multivariable Cox regression analysis revealed comorbidities including dialysis (hazard ratio [HR]: 2.10; 95% confidence interval [CI], 1.33–3.59; P = .003), chronic obstructive pulmonary disease (HR: 1.40; 95% CI, 1.01–1.77; P = .02), and immunosuppression (HR: 1.69; 95% CI, 1.17–2.43; P = .005) as hazards associated with mortality among patients with IE (Table 4). An emergent salvage operation was associated with the highest hazard for mortality when compared with an elective operation (HR: 6.55; 95% CI, 1.62–26.54; P = .009). Culture results demonstrating Streptococcus species were associated with a lower hazard for death (HR: 0.58; 95% CI, 0.35–0.97; P = .04) whereas operations consisting of combined aortic root replacement with tricuspid valve repair/replacement (HR: 2.50; 95% CI, 1.08–5.79; P = .03) or triple valve repair/replacement (HR: 2.35; 95% CI, 1.28–4.32; P = .006) were associated with an elevated risk of death. IVDU, however, was not significantly associated with mortality (HR: 1.21; 95% CI, 0.94–1.62; P = .19) (Table 4).
TABLE 4.
Multivariable Cox regression analysis for mortality among patients undergoing surgery for infective endocarditis
| Hazard ratio | 95% CI | P value | |
|---|---|---|---|
| Intravenous drug use | |||
| No | Reference | Reference | Reference |
| Yes (remote or recent) | 1.21 | 0.94–1.62 | .19 |
| Dialysis | 2.10 | 1.33–3.59 | .003 |
| Chronic obstructive pulmonary disease | 1.40 | 1.01–1.77 | .02 |
| Immunosuppression | 1.69 | 1.17–2.43 | .005 |
| Status | |||
| Elective | Reference | Reference | Reference |
| Emergent salvage | 6.55 | 1.62–26.54 | .009 |
| Previous coronary artery bypass grafting | 1.59 | 1.04–2.42 | .03 |
| Serum albumin | 0.68 | 0.55–0.84 | .003 |
| Ejection fraction | 0.98 | 0.97–0.99 | .007 |
| Endocarditis culture | |||
| Negative | Reference | Reference | Reference |
| Streptococcus | 0.58 | 0.35–0.97 | .04 |
| Operation type | |||
| Aortic valve repair/replacement | Reference | Reference | Reference |
| Aortic root + tricuspid valve repair/replacement | 2.50 | 1.08–5.79 | .03 |
| Triple valve repair/replacement | 2.35 | 1.28–4.32 | .006 |
Note: Other variables included in the multivariate model: previous myocardial infarction, New York Heart association class, congestive heart failure, peripheral vascular disease, hypertension, diabetes, median income, periannular abscess.
Abbreviation: CI, confidence interval.
4 |. CONCLUSIONS
The major finding of our study was that IVDU was not a significant predictor of mortality in patients undergoing surgery for IE, with comparable unadjusted and risk-adjusted survival rates as compared with non-IVDU. This finding can be interpreted in two different ways. As the IVDU-associated endocarditis population is often found to be noncompliant with high rates of recidivism, many clinicians believe that they are at higher risk for worse long-term outcomes. In this respect, our study is reassuring in that the majority of IVDU patients undergoing surgery for IE were alive at 5 years and had similar outcomes as non-IVDU. In contrast, some may argue that the fact that a young, otherwise healthy population is only able to achieve longer-term survival comparable to a much older, sicker patient population points to other challenges in their care unrelated to direct surgical recovery.
Prior studies have also demonstrated comparable postsurgical survival rates between IVDU and non-IVDU in the setting of IE.6,9 Our survival rates of 81.45% and 81.10% at 1-year and 57.25% and 67.64%% at 5-years for IVDU and non-IVDU were consistent with a recent meta-analysis which described 81% and 85% 1-year survival and 62.1% and 70.3% 5-year survival for IVDU and non-IVDU, respectively.5 Similarly, Kim et al6 investigated IVDU and non-IVDU patients and described survival rates of 78.9% and 76.1% at 5 years and 69.5% and 68.7% at 10 years in patients with IVDU and non-IVDU, respectively; there were no significant differences based on IVDU. Collectively, we believe these prior data in addition to our series support the notion that although acceptable survival can be achieved in patients with IVDU undergoing often complex surgical intervention for IE, there is considerable room to improve these outcomes in this young and otherwise healthy population.
Other risk factors, such as comorbidities, urgency of surgery, type of surgery, and pathogenic species were found to be independently associated with mortality in our analysis unlike IVDU which had no impact. Streptococcus species were protective from mortality when compared with a negative culture. Though not statistically significant in the model, Staphylococcus aureus, the next most commonly isolated species has been associated with a more severe clinical presentation and worse long-term survival in prior literature.10,11 Our patients with IVDU more commonly had MRSA and MSSA positive cultures. Though we found no difference in mortality, Staphylococcus aureus IE has been associated with a 57.1% increased adjusted risk of in-hospital mortality.12
As with all cardiac operations, accurate risk estimations are an important component of surgical decision-making and patient prognostication. IE can be a challenging scenario to assess risk as the operations are typically complex cases (ie, not one of the seven STS index category cases) that involve a combination of valves and reconstructive procedures. In addition, psychosocial factors including compliance and abstinence from continued or recurrent drug use are difficult to assess and capture in such risk calculations. Nonetheless, multiple prior studies have developed calculators for risk stratification in IE.13,14 An evaluation of these various risk scores found one that was developed using national STS data to be the most predictive with a c-index of 0.76.15 This particular risk index was derived in a study of over 19 000 patients with IE in the STS database which identified risk factors for 30-day mortality, including operative urgency, cardiogenic shock, preoperative renal failure, diabetes, prior cardiac surgery, chronic lung disease, and multiple valve involvement.16 Although their study was designed to identify risk factors for 30-day mortality, many of the risk factors in their score were also identified in our study as predictors of longitudinal mortality risk.
In addition to survival, it is also important to consider the impact of repeat hospitalization and recurrent endocarditis. As highlighted in Figure 2, over 50% of patients are readmitted within 5 years of surgery, with a significant proportion of these patients being readmitted within the first postoperative year. In addition, IVDU patients have higher readmission rates than non-IVDU, mostly driven by higher rates of readmission for drug-use or recurrent endocarditis. At our institution, we convene a multidisciplinary team to review all cases of recurrent IVDU-associated IE. It is important to ensure that early recurrences of IE are not related to inadequate debridement or clearance of the infection during the index case. Comparisons of more recent and prior microbiology results are therefore useful. The decision to reoperate is then individualized and based on family and social support, adequacy of addiction treatment, engagement of the patient, and operative risks of the procedure.
Recidivism is a known risk in IVDU with IE, which led the 2016 American Association for Thoracic Surgery Expert Consensus Guidelines on Endocarditis to recommend adherence to standard surgical indications with a focus on addiction treatment referral as a component of postoperative care.11 Addiction medicine consultation and initiation of medication-assisted addiction treatment, among other interventions, represent potential strategies to reduce overall mortality rates among postoperative patients with IVDU-IE.17,18 Many groups have suggested that initiation of such interventions can occur in the acute postoperative phase where many of these patients remain hospitalized either at the index surgical center or in rehabilitation centers due to the need for monitoring during their intravenous antibiotic therapy. In addition, close follow-up of all IE patients, with a multi-disciplinary management strategy particularly for IVDU patients, may assist in identifying those at highest risk for readmission thereby allowing for early recognition of complications and possibly outpatient management, where appropriate, to decrease the overall impact of IE on the health system.8 Engaging family and other social support is also essential to reducing the risks of recidivism in the IVDU subset.
This study has several limitations. Patients were identified retrospectively thus we were unable to account for all factors contributing to the decision for medical versus surgical management; this work reflects only those patients who underwent operative repair. We have performed internal quality control analyses and have found that over 90% of patients that are readmitted are readmitted within our multi-hospital regional network of 40 hospitals, however, we cannot exclude the possibility that patients were readmitted to a facility outside of our system. Finally, follow-up compliance can be poor, particularly among patients with IVDU, and the lack of a standardized follow-up schedule may also contribute to variable follow-up patterns. We tried to account for this by censoring patients at their last recorded follow-up visit.
In conclusion, we report our experience with 831 consecutive patients undergoing surgery for IE. At 5-year, the majority of patients were alive although readmission rates were high. IVDU was not a risk factor for longitudinal mortality, however, IVDU had higher readmission risk particularly for drug-use or recurrent endocarditis. We believe these data are supportive of surgical management of IE in the IVDU population when indicated, though these patients have unique challenges with regard to psychosocial outcomes and recidivism. Multi-disciplinary approaches incorporating addiction treatment are essential to IVDU-related IE care and can hopefully yield improvements in patient outcomes and reductions in healthcare resource utilization, although the optimal approach of such efforts will require further research and evaluation.
Supplementary Material
ACKNOWLEDGMENT
Dr. Arman Kilic serves on a Medical Advisory Board for Medtronic, Inc. Dr. Thomas Gleason serves on a Medical Advisory Board for Abbott, Inc. All other authors have nothing to disclose with regard to commercial support.
Footnotes
MEETING PRESENTATION
STS 56th Annual Meeting, 25 to 28 January 2020, New Orleans, LA
SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section.
REFERENCES
- 1.Toyoda N, Chikwe J, Itagaki S, Gelijns AC, Adams DH, Egorova NN. Trends in infective endocarditis in California and New York state, 1998–2013. JAMA. 2017;317:1652–1660. 10.1001/jama.2017.4287 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Desimone DC, Tleyjeh IM, Correa De Sa DD, et al. Temporal trends in infective endocarditis epidemiology from 2007 to 2013 in Olmsted County, MN. Am Heart J. 2015;170:830–836. 10.1016/j.ahj.2015.07.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schranz AJ, Fleischauer A, Chu VH, Wu LT, Rosen DL. Trends in drug use-associated infective endocarditis and heart valve surgery, 2007 to 2017: a study of statewide discharge data. Ann Intern Med. 2019;170: 31–40. 10.7326/M18-2124 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hall R, Shaughnessy M, Boll G, et al. Drug use and postoperative mortality following valve surgery for infective endocarditis: a systematic review and meta-analysis. Clin Infect Dis. 2019;69:1120–1129. 10.1093/cid/ciy1064 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Goodman-Meza D, Weiss RE, Gamboa S, et al. Long term surgical outcomes for infective endocarditis in people who inject drugs: a systematic review and meta-analysis. BMC Infect Dis. 2019;19:1–11. 10.1186/s12879-019-4558-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kim JB, Ejiofor JI, Yammine M, et al. Surgical outcomes of infective endocarditis among intravenous drug users. J Thorac Cardiovasc Surg. 2016;152:832–841. 10.1016/j.jtcvs.2016.02.072 [DOI] [PubMed] [Google Scholar]
- 7.Nguemeni Tiako MJ, Mori M, Bin Mahmood SU, et al. Recidivism is the leading cause of death among intravenous drug users who underwent cardiac surgery for infective endocarditis. Semin Thorac Cardiovasc Surg. 2019;31:40–45. 10.1053/j.semtcvs.2018.07.016 [DOI] [PubMed] [Google Scholar]
- 8.Tookes H, Diaz C, Li H, Khalid R, Doblecki-Lewis S. A cost analysis of hospitalizations for infections related to injection drug use at a county safety-net hospital in Miami, Florida. PLoS One. 2015;10:1–11. 10.1371/journal.pone.0129360 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Leahey PA, Lasalvia MT, Rosenthal ES, Karchmer AW, Rowley CF. High morbidity and mortality among patients with sentinel admission for injection drug use-related infective endocarditis. Open Forum Infect Dis. 2019;6:1–6. 10.1093/ofid/ofz089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Han SM, Sorabella RA, Vasan S, et al. Influence of Staphylococcus aureus on outcomes after valvular surgery for infective endocarditis. J Cardiothorac Surg. 2017;12:1–8. 10.1186/s13019-017-0623-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pettersson GB, Coselli JS, Pettersson GB, et al. The American Association for Thoracic Surgery (AATS) consensus guidelines: surgical treatment of infective endocarditis: executive summary. J Thorac Cardiovasc Surg. 2016;2017(153):1241–1258. 10.1016/j.jtcvs.2016.09.093 [DOI] [PubMed] [Google Scholar]
- 12.Federspiel JJ, Stearns SC, Peppercorn AF. Increasing US rates of endocarditis with Staphylococcus aureus: 1999–2008. Arch Intern Med. 2012;172:363–365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fernández-Hidalgo N, Ferreria-González I, Marsal JR, et al. A pragmatic approach for mortality prediction after surgery in infective endocarditis: optimizing and refining EuroSCORE. Clin Microbiol Infect. 2018;24: 1102.e7–1102.e15. 10.1016/j.cmi.2018.01.019 [DOI] [PubMed] [Google Scholar]
- 14.Di Mauro M, Dato GMA, Barili F, et al. A predictive model for early mortality after surgical treatment of heart valve or prosthesis infective endocarditis. The EndoSCORE. Int J Cardiol 2017;241:97–102. 10.1016/j.ijcard.2017.03.148 [DOI] [PubMed] [Google Scholar]
- 15.Varela L, López-Menéndez J, Redondo A, et al. Mortality risk prediction in infective endocarditis surgery: reliability analysis of specific scores. Eur J Cardio-Thoracic Surg. 2018;53:1049–1054. 10.1093/ejcts/ezx428 [DOI] [PubMed] [Google Scholar]
- 16.Gaca JG, Sheng S, Daneshmand MA, et al. Outcomes for endocarditis surgery in North America: a simplified risk scoring system. J Thorac Cardiovasc Surg. 2011;141:98–106. 10.1016/j.jtcvs.2010.09.016 [DOI] [PubMed] [Google Scholar]
- 17.Marks LR, Munigala S, Warren DK, Liang SY, Schwarz ES, Durkin MJ. Addiction medicine consultations reduce readmission rates for patients with serious infections from opioid use disorder. Clin Infect Dis. 2019;68:1935–1937. 10.1093/cid/ciy924 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mattick R, Kimber J, Breen C, Davoli M Buprenorphine maintenance versus placebo or methadone maintenance for opioid dependence. Cochrane Database Syst Rev 2003. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.