Highlights
-
•
First-time atrial fibrillation (AF) or heart failure (HF) before mitral valve surgery (MVS) raised the risk of post-MVS HF admission.
-
•
AF and HF showed similar increases in postdischarge HF admission risk.
-
•
Pre-MVS HF was linked to higher 30-day mortality.
-
•
Mortality after 30 days did not differ for patients with prior HF.
-
•
Pre-MVS AF was not linked to higher short- or long-term mortality.
ABSTRACT
Background
In patients with mitral regurgitation (MR), atrial fibrillation (AF), or heart failure (HF), are common indications for mitral valve surgery (MVS). However, whether first-time admissions for AF and HF are associated with similar prognostic impact remains unclear.
Aim
To evaluate the risk of mortality after MVS among patients with recent hospitalization for AF or HF.
Methods
Using Danish nationwide registries (2000-2023), we identified patients undergoing first-time MVS for MR. Based on admissions within 6 months before surgery, patients were categorized as: (1) no new AF/HF (no first-time admission for AF or HF), (2) AF group (first-time admission for AF but not HF), or (3) HF group (first-time admission for HF, irrespective of admission for AF). We examined the 1-year rate of HF admission from discharge using multivariable Cox regression. From the date of MVS, we examined the rate of mortality.
Results
Compared with patients with no new AF/HF, the adjusted rate of admission for HF was significantly higher in the AF group (hazard ratio [HR] 1.55; 95% CI: 1.19-2.02) and the HF group (HR 1.67; 95% CI: 1.33-2.11). When examining mortality, we observed a higher 30-day risk among patients in the HF group, and no difference between groups beyond 30 days of follow-up.
Conclusion
In patients undergoing MVS for MR, first-time admission for AF or HF within 6 months before MVS was associated with increased risk of admission for HF after MVS. Mortality appeared higher in the HF group during the initial 30 days postsurgery.
Graphical abstract
Background
Mitral regurgitation (MR) is the second most common reason for heart valve surgery in the Western World and one of the most common valvular heart disease. Several studies have demonstrated favorable outcomes following mitral valve surgery (MVS) ,,, ; however, current risk stratification among patients undergoing surgery for MR relies primarily on generic cardiothoracic risk models, and more evidence on prognostic markers is needed to identify high-risk patient groups.
Patients with severe chronic primary MR often remain asymptomatic for years owing to compensatory pathophysiologic mechanisms. Once these mechanisms fail, complications such as heart failure (HF) and atrial fibrillation (AF) may develop, prompting surgery in both symptomatic and asymptomatic patients.
Among patients with severe primary MR, HF is a class I indication for MVS and is associated with increased risk of postoperative mortality and reduced left ventricular ejection fraction. ,, Likewise, observational studies have demonstrated that AF is associated with worse outcomes, regardless of surgical intervention. ,,,,
HF and AF are thus considered important risk factors in MR, and first-time admission for one of these conditions may be interpreted as a proxy for progression in disease severity. However, the prognostic significance of first-time hospital admissions for these conditions before surgery remains uncertain, and no studies have examined the difference in the associated prognostic impact of these two comorbidities separately. Clarifying this difference in associated risk may improve surgical risk stratification and enhance patient counseling. Therefore, this study examined the 1-year outcomes of patients with MR who had a first-time admission for HF or AF before undergoing MVS.
Methods
Data sources
All Danish citizens are assigned a unique personal registration number, which enables the linkage between administrative Danish registries and clinical registries on a nationwide basis. Tax-financed health care is provided to all Danish citizens. This study used the following registries: (1) The Danish National Patient Registry (DNPR), which holds records on all hospital admissions since 1977 and outpatient visits since 1995, with discharge diagnoses based on the International Classification of Diseases (ICD) (ICD-8 and ICD-10 codes) and surgical procedures classified according to The Nordic Medico-Statistical Committee since 1996 ; (2) The Civil Registration System, which holds information on sex, vital-status, migration, and date of birth ; and (3) The Danish National Prescription Registry, which holds records on all claimed drug prescriptions since 1995.
Study population
Figure 1 shows the patient selection process. The study population comprised all adult Danish residents admitted for first-time MVS. Patients were excluded according to the following exclusion criteria: (1) transcatheter edge-to-edge repair; (2) no MR diagnosis before or during the index admission; (3) age < 18 years; (4) admission or discharge outside the study period of January 1, 2000 to December 31, 2023; (5) diagnosis of infective endocarditis during the index admission; and (6) diagnoses of myocardial infarction or related complications during index admission to ensure that the MR diagnosis was not related to acute ischemic complications. Notably, patients receiving concomitant surgical procedures such as coronary artery bypass graft (CABG) or aortic valve intervention were included in the main population.
Selection of study population.
This flow chart illustrates the population selection process.
Exposure
Patients were categorized into three groups: (1) no new AF/HF; (2) AF group; and (3) HF group. Patients in the no new AF/HF group were defined as those without a first-time admission for either AF or HF within 6 months before MVS admission. Notably, these patients could still have had hospital admissions for AF or HF occurring more than 6 months before MVS admission, or they could have received an outpatient diagnosis of AF or HF during the 6-month preoperative window. The AF group included patients with a first-time admission for AF, but not HF, within the 6-month period. The HF group included patients with a first-time admission for HF within 6 months before MVS admission; this group could also include individuals with a concurrent diagnosis code of first-time AF during the same time period. AF and HF were identified using the primary or secondary ICD-10 codes I48 and I50, respectively. The overall validity of cardiovascular diagnoses in the DNPR is high, with the HF diagnosis code having been validated with a positive predictive value of 76% to 81 while the AF diagnosis code has been validated with a positive predictive value of 95%. ,
Outcomes and follow-up
The primary outcome was hospital admission for HF (ICD-10: I50) within 1 year following discharge. For this outcome, patients were followed from the date of discharge until the date of occurrence of the outcome, death, end of the study period (December 31, 2023), or a maximum of 1 year of follow-up, whichever came first. The secondary outcome was all-cause mortality which was assessed within 1 year after MVS. Follow-up for mortality was further stratified into two periods: from 0 to 30 days after MVS and from 30 to 365 days after MVS. Follow-up was restricted to 1 year to minimize competing risks and unrelated comorbidities (eg, age-related HF progression and new ischemic disease).
Covariates
Patients’ medical history was obtained from the DNPR using inpatient and outpatient diagnosis codes at any time before the date of the index admission. Pharmacotherapy was defined as claimed prescriptions within 6 months before admission. Supplementary Tables I and II list all the classification codes used for medical history and pharmacotherapy, respectively.
Statistics
Patient characteristics were compared between study groups. Categorical variables were expressed as counts and percentages, with differences assessed using the χ 2 test. Continuous variables were presented as medians with interquartile ranges (25th-75th percentiles), and differences were assessed using the Wilcoxon rank-sum test. The absolute risk of admission with HF was examined using the Aalen-Johanson estimator, considering death as a competing risk, and statistical differences between cumulative incidence curves were assessed using Gray’s test. The absolute risk of mortality was assessed using the Kaplan-Meier estimator, and differences between survival curves were assessed using the log-rank test. To examine the adjusted rates of outcomes, hazard ratios (HR) with 95% confidence intervals (CI) were calculated using multivariate Cox proportional hazards regression. Cox regression models were adjusted for sex, age, calendar period (2000-2007; 2008-2015; and 2016-2023), ischemic heart disease, malignancy, chronic obstructive pulmonary disease, liver disease, peripheral vascular disease, diabetes, ischemic or hemorrhagic stroke, chronic kidney disease, and concomitant surgery with CABG or aortic procedures. The assumption of proportional hazards was tested and fulfilled.
All statistical analyses were performed using SAS statistical software (version 9.4, Cary, NC, USA) and R (version 2023.02.2). The level of statistical significance was defined as a P -value <.05.
Supplementary analyses
Four supplementary analyses were conducted to assess the robustness of our findings. First, a fourth exposure group was added, comprising patients who had a first-time admission for both AF and HF within 6 months before surgery. Second, outcomes were re-evaluated among patients undergoing isolated mitral valve repair, ie, excluding those with prosthetic implantation or concomitant CABG or aortic procedures. Third, we broadened the exposure definition so that patients were classified as exposed if they had any admission for AF or HF within the specified 6-month time window, regardless of whether this was a first-time admission or a recurrent admission. Finally, outcomes were re-examined using exposure definitions based on both first-time admissions and outpatient diagnoses. In the supplementary analyses, we used the same multivariable models as in the primary analyses, including the identical set of covariates; only the exposure groups were changed.
Ethics
In Denmark, registry-based studies conducted for the sole purpose of statistics and scientific research do not require ethical approval or informed consent by law. The study was approved by the data responsible institute (Capital Region of Denmark, Approval number: P-2019-191) in accordance with the General Data Protection Regulation (GDPR).
Results
Study population and baseline characteristics
Table displays patients’ medical history before admission for MVS. From January 1, 2000, to December 31, 2023, we identified 6,774 adult patients undergoing first-time MVS for MR. Among these, 5,689 (84.0%) were in the no new AF/HF group (67.2% male, median aged 67 years), 516 (7.6%) in the AF group (71.0% male, median aged 68 years), and 569 (8.4%) in the HF group (65.4% male, median aged 68 years). Consistent with the younger age of patients without new AF/HF, we observed a stepwise increase in comorbidity burden from the no–new AF/HF group to the AF group and further to the HF group. Mitral valve repair was performed in 70.5% of patients with no new AF/HF, 73.1% of the AF group, and 63.3% of the HF group. Rates of concomitant aortic valve intervention were similar across groups (11%-16%). Concomitant AF intervention was more frequent in the AF group (51.2%) compared to the no new AF/HF (19.3%) and HF (22.3%) groups. Conversely, the HF group more frequently underwent concomitant CABG (38.1%), compared to 23.6% in the no new AF/HF group and 19.6% in the AF group. Supplementary Table III displays baseline characteristics among those surviving admission for MVS.
Table
Baseline characteristics.
|
No new AF/HF
N = 5689 |
AF group
N = 516 |
HF group
N = 569 |
P -value | |
|---|---|---|---|---|
| Male sex, N (%) | 3666 (67.2) | 353 (71.0) | 344 (65.4) | .21 |
| Median age, years (p25-p75) |
66.5
(57.4-73.7) |
68.2
(60.6-74.1) |
68.2
(58.8-74.6) |
<.01 |
| Period, N (%) | <.01 | |||
| 2000-2007 | 1613 (28.4) | 131 (25.4) | 303 (53.3) | – |
| 2008-2015 | 1994 (35.1) | 175 (33.9) | 167 (29.4) | – |
| 2016-2023 | 2082 (36.6) | 210 (40.7) | 99 (17.4) | – |
| Median days from MR diagnosis to MVS (p25-p75) | 98 (27-604) | 77 (40-174) | 74 (33-158) | <.01 |
| Type of mitral valve surgery, N (%) | ||||
| Repair | 4013 (70.5) | 377 (73.1) | 360 (63.3) | <.01 |
| Mechanical valve | 705 (12.4) | 62 (12.0) | 102 (17.9) | – |
| Bioprosthetic valve | 971 (17.1) | 77 (14.9) | 107 (18.8) | – |
| Procedures performed concomitant with MVS, N (%) | ||||
| Tricuspid valve intervention | 629 (11.1) | 79 (15.3) | 68 (12.0) | .88 |
| Aortic valve intervention | 702 (12.3) | 60 (11.6) | 94 (16.5) | .37 |
| Coronary artery bypass graft surgery | 1344 (23.6) | 101 (19.6) | 217 (38.1) | <.01 |
| Surgical ablation | 1022 (18.0) | 237 (45.9) | 120 (21.0) | <.01 |
| Left atrial appendage intervention | 353 (6.2) | 85 (16.5) | 27 (4.8) | <.01 |
| Length of stay, days (p25-p75) | 11 (8-19) | 11 (8-18) | 14 (9-25) | <.01 |
| First-time admission within 6 mo before MVS admission, N (%) | ||||
| Heart failure | 0 (-) | 0 (-) | 526 (100.0) | N/A |
| Atrial fibrillation | 0 (-) | 516 (100.0) | 143 (25.1) | N/A |
| Diagnoses given at any time before the index admission, N (%) | ||||
| History of heart failure | 1233 (21.7) | 88 (17.1) | 569 (100.0) | <.01 |
| History of atrial fibrillation | 1944 (34.2) | 516 (100.0) | 280 (49.2) | <.01 |
| Ischemic heart disease | 1889 (33.2) | 137 (26.6) | 280 (49.2) | <.01 |
| Coronary artery bypass graft surgery | 77 (1.4) | 9 (1.7) | 7 (1.2) | .73 |
| Percutaneous coronary intervention | 253 (4.5) | 21 (4.1) | 33 (5.8) | .29 |
| Permanent pacemaker | 126 (2.2) | 11 (2.1) | 17 (3.0) | .49 |
| Chronic obstructive pulmonary disease | 425 (7.5) | 36 (7.0) | 77 (13.5) | <.01 |
| Chronic kidney disease | 263 (4.6) | 36 (7.0) | 43 (7.6) | <.01 |
| Peripheral artery disease | 177 (3.1) | 17 (3.3) | 29 (5.1) | .04 |
| Ischemic/hemorrhagic stroke | 346 (6.1) | 22 (4.3) | 46 (8.1) | .03 |
| Diabetes | 332 (5.8) | 25 (4.8) | 55 (9.7) | <.01 |
| Liver disease | 117 (2.1) | 11 (2.1) | 21 (3.7) | .04 |
| Redeemed prescription within 6 mo before MVS admission, N (%) | ||||
| Beta blockers | 1958 (34.4) | 380 (73.6) | 314 (55.2) | <.01 |
| Statins | 1914 (33.6) | 169 (32.8) | 209 (36.7) | .29 |
| Calcium channel blockers | 940 (16.5) | 90 (17.4) | 108 (19.0) | .30 |
| Oral anticoagulants | 1552 (27.3) | 418 (81.0) | 243 (42.7) | <.01 |
| RAS inhibitors | 2430 (42.7) | 220 (42.6) | 373 (65.6) | <.01 |
| Mineralocorticoid receptor antagonists | 406 (7.1) | 29 (5.6) | 149 (26.2) | <.01 |
| Loop diuretics | 2274 (40.0) | 335 (64.9) | 489 (85.9) | <.01 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
