Abstract
Background
Although major non-cardiac surgery is common, few large-scale studies have examined the incidence and consequences of post-operative atrial fibrillation (POAF) in this population. We sought to define the incidence of POAF and its impact on outcomes after major non-cardiac surgery.
Methods
Using administrative data, we retrospectively reviewed the hospital course of adults who underwent major non-cardiac surgery at 375 US hospitals over a 1-year period. Clinically significant POAF was defined as atrial fibrillation occurring during hospitalization that necessitated therapy.
Results
Of 370447 patients, 10957 (3.0%) developed clinically significant POAF while hospitalized. Of patients with POAF, 7355 (67%) appeared to have pre-existing atrial fibrillation and 3602 (33%) had newly diagnosed atrial fibrillation. Black patients had a lower risk of POAF (Adjusted Odds Ratio, 0.53; 95% CI, 0.48 to 0.59; P<0.001). Patients with POAF had higher mortality (Adjusted Odds Ratio, 1.72; 95% CI, 1.59 to 1.86; P<0.001), markedly longer length of stay (Adjusted Relative Difference, +24.0%; 95% Confidence Interval [CI], +21.5% to +26.5%; P<0.001), and higher costs (Adjusted Difference, +$4177; 95% CI, +$3764 to + $4590; P<0.001). These findings did not differ by whether POAF was a recurrence of pre-existing atrial fibrillation, or a new diagnosis.
Conclusion
POAF following non-cardiac surgery is not uncommon and is associated with increased mortality and cost. Our study identifies risk factors for POAF, which appear to include race. Strategies are needed to not only prevent new POAF, but also improve management of patients with pre-existing atrial fibrillation.
Keywords: atrial fibrillation, non-cardiac surgery, outcomes, complications
INTRODUCTION
Atrial fibrillation affects approximately 2.5 million people in the United States and roughly 10% of people over 80 years of age.1 This dysrhythmia has been shown to be associated with increased mortality, even in young relatively healthy people.2 Patients that develop atrial fibrillation in the peri-operative setting represent a poorly characterized subset of this population. An estimated 40–50 million non-cardiac surgeries are performed in this country each year and the number of procedures, as well as the average age of the patients, is steadily rising. As a result, the incidence of post-operative atrial fibrillation (POAF) can be expected to increase over time.
Recent evidence suggests that POAF complicating cardiac surgery increases health care costs and precipitates subsequent neurocognitive decline. As a result, strategies have been actively sought for the prevention of POAF after cardiac surgery. Subsequent research has shown that peri-operative administration of amiodarone, beta-blockers, and statins effectively decrease POAF in this population.7–10
The risks, mechanisms, and sequelae of atrial fibrillation after cardiac surgery are, at least in part, linked to the underlying cardiac disease and direct manipulation of the pericardium and myocardium during surgery, and therefore should not be fully generalized to patients undergoing non-cardiac surgery. Relatively little is known about the POAF as a complication of non-cardiac surgery. Pre-existing atrial fibrillation has been shown to be independently associated with increased mortality in patients undergoing non-cardiac surgery.11 New-onset POAF following non-cardiac surgery has been studied only in small single-center studies. Given the large number of people who undergo these types of procedures, defining the scope of POAF in this population is important. We therefore assessed the incidence of POAF and its impact on outcomes after major non-cardiac surgery in a large nationwide cohort.
METHODS
Design Overview
We performed a retrospective cohort study using data from Premier Perspective, a database developed for quality and utilization benchmarking by Premier Incorporated, Charlotte, NC. The methods and design are the same as in our recently published data from this registry.14 In addition to data elements available in the standard hospital discharge file, the Perspective database contains a date-stamped log at the individual patient level of all billed items, including medications as well as laboratory, diagnostic, and therapeutic services.
Setting and Participants
Patients were included in the analysis if they were 18 years of age or older and underwent major non-cardiac surgery between January 1, 2008 and December 31, 2008 at any of the 375 hospitals participating in Premier Perspective. Surgical procedures were categorized based on Diagnosis Related Groups (using APR-DRG software, version 27.0, 3M Corp, Minneapolis, MN) and were considered major if the mean length of stay for patients in the diagnosis related group was 3 days or more.15 Procedures classified as cardiac and obstetric were excluded. The included list of procedures was then organized into nine categories of surgical type.
In order to discern new atrial fibrillation from chronic atrial fibrillation, cohort selection was focused on patients whose data had accurate and complete present-on-admission coding. These codes represent a date stamp in administrative data for secondary diagnoses and help determine which diagnoses are present at the time of admission. 16–18 Present-on-admission codes have been validated for in-hospital diagnoses of atrial fibrillation.19 For any given hospital, we included patients who underwent surgery beginning the month that 100% of the cases of atrial fibrillation in that hospital had complete present-on-admission coding. Three hundred and seventy-five hospitals submitted data to Premier during the study time period. Of this 375, 46 hospitals were excluded because they had incomplete present-on-admission coding for all 12 months of the study period.
For each case, patient demographic information and hospital characteristics were recorded. Co-morbidities were established using a combination of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) secondary diagnoses using standard methods.20
Permission to perform the study was obtained by the institutional review board at UCSF, San Francisco, CA and the requirement for informed consent was waived due to minimal risk to patients.
Outcomes and Follow-up
Definition of Post-Operative Atrial Fibrillation (POAF)
The designation of significant post-operative atrial fibrillation (POAF) was made using a secondary diagnosis ICD-9-CM code for atrial fibrillation (427.31) in combination with charge or procedure code data that suggested that the patient’s dysrhythmia necessitated therapy. A combination of an ICD-9-CM code of 427.31 and one of the following was required for a diagnosis of POAF: an ICD-9-CM procedure code for electrical cardioversion in the post-operative period; pharmacy charge data for a new prescription of digoxin in the post-operative period; or pharmacy charge data for intravenous beta-blockers, intravenous calcium channel blockers, or intravenous amiodarone in the post-operative period.
Definition of peri-operative use of medications
We used pharmacy records to identify whether certain classes of medications were administered during the peri-operative period. We limited our “peri-operative” time window to day 0 (the day of surgery) and day 1.
Definition of agents likely to affect POAF risk
Peri-operative administration of digoxin, beta-blockers, angiotensin-converting enzyme inhibitors (ACE-Inhibitor), angiotensin receptor blockers (ARB), calcium channel blockers, anti-platelet, agents, or anti-arrhythmic drugs were assessed using pharmacy records. A similar approach was used to assess post-operative use of inotropes, vasopressors, warfarin, digoxin, intravenous beta-blockers, intravenous calcium channel blockers, or anti-arrhythmic drugs.
Administrative Outcomes
In-hospital mortality, length of stay, actual hospitalization costs, and 15-day and 30-day readmission rates were obtained from the Perspective discharge file.
Statistical Analysis
Summary statistics for the overall sample were constructed by using frequencies and proportions for categorical data and mean, median, and inter-quartile ranges for continuous variables. We also stratified these summary statistics by whether patients developed POAF.
Univariate logistic regression was used to determine the unadjusted association between patient, hospital, and surgery characteristics and POAF. We then constructed a multivariate logistic regression model to examine the association (adjusted odds ratio) between the covariates and POAF. A generalized estimation equations model with logit link, independence working correlation, and robust standard errors was employed. Two methods were used to determine which covariates to include in the model. First, a group of likely confounders (age, for example) was specified a priori and automatically included in the model for face validity. Second, another list of possible confounders was generated from clinical knowledge and data from prior studies. Confounders from this second group were kept in the model if they were found to change the model in a statistically significant manner (p <0.05). Interaction terms chosen for clinical or biological plausibility were added to the base model and tested for statistical significance (p <0.05). Co-variates were only allowed into the model if they were known at the time of surgery (i.e. post-operative medications or complications were not candidate co-variates). Variables included in the final model included: age, gender, race, hospital size (number of beds), surgery type, admission source, peri-operative administration of certain medications (statins, anti-arrhythmic drugs, beta-blockers, ACE-Inhibitors, ARBs, digoxin), and selected co-morbidities (hypertension, diabetes, congestive heart failure, ischemic heart disease, renal insufficiency, cerebrovascular disease, alcohol abuse, obesity, and pulmonary vascular disease).
Using the same methodology, models assessing POAF as the predictor of interest were constructed with multivariate logistic regression for dichotomous outcomes (mortality and readmission) and linear regression for continuous outcomes (cost and length of stay). In order to limit skew and maintain compliance with model assumptions, the cost variable was truncated (by dropping the upper 5% of the tail) and the length of stay variable was log transformed. Robust standard errors were used to account for clustering by hospital, arising from unmeasured differences in practice and expertise.
All analyses were performed using STATA version 10 (Stata Corporation, College Station, Texas). No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the paper and its final contents.
RESULTS
Patient characteristics (Tables I, II, III)
Table I.
Clinical Characteristics of Patients Undergoing Major Non-Cardiac Surgery, Stratified by the Development of POAF
| Characteristics | POAF N = 10957 (3.0%) |
No POAF N = 359490 (97.0%) |
P value |
|---|---|---|---|
| Age (years), mean (SD) | 74.6 (10.6) | 62.4 (16.2) | <0.001 |
| Male gender | 5435 (49.6%) | 153831 (42.8%) | <0.001 |
| Race/ethnicity | <0.001 | ||
| White | 8337 (76.1%) | 252376 (70.2%) | |
| Black | 628 (5.7%) | 39004 (10.9%) | |
| Hispanic | 226 (2.1%) | 12905 (3.6%) | |
| Other | 1766 (16.1%) | 55205 (15.4%) | |
| Admission Source | <0.001 | ||
| Emergency Room | 4940 (45.1%) | 112728 (31.4%) | |
| Inter-facility transfer | 871 (7.9%) | 15096 (4.2%) | |
| Outpatient referral | 4923 (44.9%) | 223831 (62.3%) | |
| Other | 223 (2.0%) | 7835 (2.2%) | |
| Co-morbidities | |||
| Congestive Heart Failure | 3340 (30.5%) | 25319 (7.0%) | <0.001 |
| Valvular disease | 1720 (15.7%) | 16345 (4.6%) | <0.001 |
| Ischemic heart disease | 4051 (37.0%) | 63210 (17.6%) | <0.001 |
| Peripheral vascular disease | 1405 (12.8%) | 23830 (6.6%) | <0.001 |
| Hypertension | 7706 (70.3%) | 204597 (56.9%) | <0.001 |
| Cerebro-vascular disease | 480 (4.4%) | 5547 (1.5%) | <0.001 |
| Chronic pulmonary disease | 3223 (29.4%) | 61986 (17.2%) | <0.001 |
| Diabetes | 3269 (29.8%) | 80236 (22.3%) | <0.001 |
| Dyslipidemia | 3741 (34.1%) | 101698 (28.3%) | <0.001 |
| Hypothyroidism | 1789 (16.3%) | 43473 (12.1%) | <0.001 |
| Renal failure | 2094 (19.1%) | 35846 (10.0%) | <0.001 |
| Coagulopathy | 1333 (12.2%) | 13769 (3.8%) | <0.001 |
| Obesity | 1193 (10.9%) | 44558 (12.4%) | <0.001 |
| Fluid & electrolyte disorders | 5037 (46.0%) | 70594 (19.6%) | <0.001 |
| Anemia | 3995 (36.5%) | 81961 (22.8%) | <0.001 |
| Alcohol abuse | 366 (3.3%) | 10261 (2.9%) | 0.003 |
Abbreviations: POAF, Post-operative atrial fibrillation
Table II.
Surgery Type and Medications Administered during Hospitalization, Stratified by the Development of POAF
| Characteristics | POAF N = 10957 (3.0%) |
No POAF N = 359490 (97.0%) |
P value |
|---|---|---|---|
| Type of surgery | <0.001 | ||
| Intracranial | 326 (3.0%) | 8327 (2.3%) | |
| Spinal/Peripheral Nerve | 397 (3.6%) | 31113 (8.7%) | |
| Abdominal | 3895 (35.6%) | 86715 (24.1%) | |
| Genital-urinary | 486 (4.4%) | 26678 (7.4%) | |
| Orthopedic | 2425 (22.1%) | 136725 (38.0%) | |
| Otolaryngologic | 347 (3.2%) | 5279 (1.5%) | |
| Thoracic (Non-cardiac) | 465 (4.2%) | 7582 (2.1%) | |
| Vascular | 570 (5.2%) | 20590 (5.7%) | |
| Miscellaneous | 2046 (18.7%) | 36481 (10.2%) | |
| Peri-Operative Medications | |||
| Statins | 2114 (19.3%) | 77757 (21.6%) | <0.001 |
| Beta-blockers | 3585 (32.7%) | 100575 (28.0%) | <0.001 |
| ACE inhibitors | 2143 (19.6%) | 71323 (19.8%) | 0.466 |
| Angiotensin receptor blockers | 817 (7.5%) | 35416 (9.9%) | <0.001 |
| Anti-arrhythmics | 583 (5.3%) | 4082 (1.1%) | <0.001 |
| Aspirin | 1840 (16.8%) | 48075 (13.4%) | <0.001 |
| Calcium channel blockers | 1140 (10.4%) | 41965 (11.7%) | <0.001 |
| Clopidogrel | 454 (4.1%) | 13114 (3.7%) | 0.007 |
| Prophylactic antibiotics | 9981 (91.1%) | 327752 (91.2%) | 0.774 |
| Pharmacologic VTE prophylaxis | 7103 (64.8%) | 158443 (44.1%) | <0.001 |
| Mechanical VTE prophylaxis | 2232 (20.4%) | 33188 (9.2%) | <0.001 |
| Post-operative therapy received | |||
| IV Amiodarone | 3083 (28.1%) | 1566 (0.4%) | <0.001 |
| IV beta-blocker | 3970 (36.5%) | 14547 (4.1%) | <0.001 |
| IV calcium channel blocker | 2555 (23.3%) | 1335 (0.4%) | <0.001 |
| Digoxin | 5600 (51.1%) | 3192 (0.9%) | <0.001 |
| DC Cardioversion | 810 (7.4%) | 1317 (0.4%) | <0.001 |
Abbreviations: POAF, Post-operative atrial fibrillation; ACE, Angiotensin converting enzyme; VTE, venous thrombo-embolism; IV, intravenous
Table III.
Administrative Characteristics of Patients Undergoing Major Non-cardiac Surgery and Selected Outcomes
| Characteristics | POAF N = 10957 (3.0%) |
No POAF N = 359490 (97.0%) |
P value |
|---|---|---|---|
| Primary payer | <0.001 | ||
| Uninsured | 17 (0.2%) | 2477 (0.7%) | |
| Indemnity | 387 (3.5%) | 25129 (7.0%) | |
| Managed care | 1089 (9.9%) | 97131 (27.0%) | |
| Medicare/Medicaid | 9128 (83.3%) | 206020 (57.3%) | |
| Other | 336 (3.1%) | 28733 (8.0%) | |
| Characteristics of Hospitals | |||
| Location | 0.76 | ||
| Rural | 1133 (10.3%) | 37491 (10.4%) | |
| Urban | 9824 (89.7%) | 321999 (89.6%) | |
| Area | <0.001 | ||
| Midwest | 1845 (16.8%) | 68513 (19.1%) | |
| Northeast | 3139 (28.7%) | 97051 (27.0%) | |
| South | 3702 (33.8%) | 122238 (34.0%) | |
| West | 2271 (20.7%) | 71688 (19.9%) | |
| Number of beds | <0.001 | ||
| 0–99 | 274 (2.5%) | 10888 (3.0%) | |
| 100–199 | 980 (8.9%) | 34332 (9.6%) | |
| 200–299 | 1476 (13.5%) | 50352 (14.0%) | |
| 300–499 | 4133 (37.7%) | 133877 (37.2%) | |
| > 500 | 4094 (37.4%) | 130041 (36.2%) | |
| Teaching Hospital | 2776 (25.3%) | 92407 (25.7%) | 0.38 |
| Disposition | <0.001 | ||
| Home | 2176 (19.9%) | 165734 (46.1%) | |
| Skilled Nursing Facility | 3693 (33.7%) | 71520 (19.9%) | |
| Home health care | 1988 (18.1%) | 80201 (22.3%) | |
| Died in Hospital | 1549 (14.1%) | 7402 (2.1%) | |
| Rehabilitation | 960 (8.8%) | 25949 (7.2%) | |
| Other | 591 (5.4%) | 8684 (2.4%) | |
| Outcomes | |||
| Mortality | 1549 (14.1%) | 7402 (2.1%) | <0.001 |
| Readmission 15-days | 1066 (9.7%) | 23626 (6.6%) | <0.001 |
| Readmission 30-days | 1584 (14.5%) | 35421 (9.9%) | <0.001 |
| Length of Stay Mean (SD) | 15.0 (14.0) | 6.9 (8.3) | <0.001 |
| Hospitalization Cost Mean (SD) | 42930 (49327) | 21215 (24288) | <0.001 |
Abbreviations: POAF, Post-operative atrial fibrillation
A total of 370447 patients meeting eligibility criteria underwent major non-cardiac surgery during the study period. Of these, 10957 patients (3.0%) developed POAF and 3602 patients (33% of all POAF patients) had atrial fibrillation classified as new by present-on-admission coding. Patients developing POAF were older; more often male; and much more likely to have congestive heart failure, ischemic heart disease, and hypertension.
Associations of patient race, co-morbidities, medications, and surgery type with incidence of POAF (Table IV)
Table IV.
Predictors of POAF – Race, Co-morbidities, Medications, and Surgery Type
| Overall (N=370447) | Unadjusted POAF Rate (n=10957, 3.0%) | Adjusted OR [95% CI] | P value |
|---|---|---|---|
| Black race (39632) | 628 (1.6%) | 0.53 [0.49–0.58] | <0.001 |
| Co-morbidities | |||
| CHF (28659) | 3340 (11.7%) | 2.14 [2.01–2.27] | <0.001 |
| DM (83515) | 3269 (3.9%) | 1.01 [0.96–1.06] | 0.700 |
| IHD (67261) | 4051 (6.0%) | 1.22 [1.16–1.27] | <0.001 |
| HTN (212303) | 7706 (3.6%) | 1.16 [1.11–1.22] | <0.001 |
| Medications | |||
| Beta-blocker (104160) | 3585 (3.4%) | 1.01 [0.95–1.07] | 0.678 |
| Statin (79871) | 2114 (2.6%) | 0.79 [0.72–0.86] | <0.001 |
| ACE-I (73466) | 2143 (2.9%) | 0.85 [0.81–0.90] | <0.001 |
| ARB (36233) | 817 (2.3%) | 0.73 [0.66–0.81] | <0.001 |
| Surgery Type | |||
| Orthopedic (139150) | 2425 (1.7%) | reference | reference |
| Otolaryngologic (5626) | 347 (6.2%) | 0.96 [.84–1.09] | 0.528 |
| Thoracic (8047) | 465 (5.8%) | 1.52 [1.36–1.70] | <0.001 |
| Intra-cranial (8653) | 326 (3.8%) | 1.45 [1.28–1.64] | <0.001 |
| Abdominal (90610) | 3895 (4.3%) | 1.82 [1.72–1.93] | <0.001 |
| Genito-urinary (27164) | 2046 (5.3%) | 1.08 [0.97–1.19] | 0.158 |
| Vascular (21160) | 570 (2.7%) | 0.78 [0.71–0.87] | <0.001 |
| Spine/Peripheral Nerve (31510) | 397 (1.3%) | 0.85 [0.76–0.94] | 0.003 |
| Miscellaneous (38527) | 2046 (5.3%) | 1.41 [1.31–1.51] | <0.001 |
Abbreviations: POAF, Post-operative atrial fibrillation; CHF, congestive heart failure; DM, diabetes mellitus; IHD, ischemic heart disease; HTN, hypertension; CRI, chronic renal insufficiency; ACE-I, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker
After adjustment for patient risk factors, surgery type, and hospital factors in multivariable clustered logistic models, black race was associated with a markedly lower adjusted odds of developing POAF. Congestive heart failure, ischemic heart disease, and hypertension were associated with higher adjusted odds of developing POAF. Among key medications administered in the peri-operative period, statins, ACE-Inhibitors, and ARBs were all found to be associated with a significantly lower adjusted odds of developing POAF.
In a multivariable logistic model, we examined the association of surgery type with the incidence of POAF. The largest sub-group, orthopedic procedures, was used as the reference; the crude incidence of POAF for patients undergoing orthopedic surgery was 1.7%. Abdominal surgery was associated with the highest adjusted odds for developing POAF (Adjusted Odds Ratio [AOR], 1.82; 95% Confidence Interval [CI], 1.72–1.93).
Association between POAF and outcomes (Tables V,VI)
Table V.
Outcomes, Stratified by the Development of POAF
| Outcomes | POAF (N=10957) | No POAF (N=359490) | Adjusted OR* [95% CI] | P value |
|---|---|---|---|---|
| Mortality | 1549 (14.1%) | 7402 (2.1%) | 1.72 [1.59–1.86] | <0.001 |
| Readmission at 30 days | 1586 (14.4%) | 35419 (9.8%) | 1.00 [0.94–1.06] | 0.999 |
| Adjusted Difference* [95% CI] | ||||
| Length of Stay, Median, [IQR]† | 11 [6–19] | 4 [3–8] | +24.0% [21.5%–26.5%] | <0.001 |
| Hospitalization Cost, Median [IQR]‡ | 27164 [17133–49745] | 15076 [10519–22769] | +4177 [3764–4590] | <0.001 |
Abbreviations: POAF, Post-operative atrial fibrillation
Adjusted odds ratio or adjusted difference for the outcome, associated with POAF; regression models are fully adjusted
Length of stay in days, log transformed in adjusted model
top 5th percentile of cost removed to reduce skewness, cost is in dollars
Table VI.
Outcomes, Stratified by the Development of POAF among Patients without Pre-existing Atrial Fibrillation*
| Outcomes | POAF (N=3602) | No POAF (N=339003) | Unadjusted OR [95% CI] | P value | Adjusted OR [95% CI] | P value |
|---|---|---|---|---|---|---|
| Mortality | 570 (15.8%) | 6646 (2.0%) | 9.40 [8.57–10.31] | <0.001 | 1.68 [1.52–1.86] | <0.001 |
| Readmission 30 days | 529 (14.7%) | 32338 (9.5%) | 1.63 [1.49–1.79] | <0.001 | 1.02 [0.92–1.14] | 0.673 |
| Unadjusted Difference in Means [95% CI] | Adjusted Difference in Means [95% CI] | |||||
| Length of Stay, median, [IQR]† | 13 [8–22] | 4 [3–7] | +275% [268%–182%] | <0.001 | + 37.3% [33.9%–40.9%] | <0.001 |
| Hospitalization Cost, median, [IQR]‡ | 25205 [18041–36280] | +14525 [10202–20987] | +10617 [10224–11011] | <0.001 | + 5914 [5416–6412] | <0.001 |
Abbreviations: POAF, Post-operative atrial fibrillation
Patients with pre-existing atrial fibrillation, as determined by present-on-admission coding, were excluded from this analysis
Length of stay log transformed in adjusted and unadjusted models, length of stay is in days
top 5th percentile of cost removed to reduce skewness, cost is in dollars
Outcomes were compared between patients who did and did not develop POAF. In a multivariate logistic regression model, patients developing POAF had a higher odds of mortality (AOR, 1.72; 95% CI, 1.59–1.86), longer hospital stays (Adjusted Relative Difference, +24.0%; 95% CI, 21.5%–26.5%), and increased hospitalization costs (Adjusted Difference, +$4177; 95% CI, $3764–$4590). There was no significant difference between the groups in the 30-day rate of readmission.
Analyses examining the association of new POAF with outcomes excluded all patients with pre-existing atrial fibrillation. New POAF (as determined by present-on-admission codes) had a similar association to mortality (AOR, 1.68; 95% CI, 1.52–1.86) as all POAF. New POAF was associated with extended hospital stays (Adjusted Relative Difference, +37.3%; 95% CI 33.9%–40.9%) and higher cost (Adjusted Difference, +$5914; 95% CI $5416–$6412).
A sensitivity analysis was performed by altering the outcome variable such that POAF was based solely on an ICD-9 code (427.31) with no clinical qualifiers. The association of POAF with mortality was attenuated but remained highly statistically significant (AOR, 1.24; 95% CI, 1.17–1.32; P<0.001).
COMMENT
In this large cohort of surgical patients, POAF occurred in about 3.0% of patients after non-cardiac surgery. Advancing age and congestive heart failure were associated with higher risk for POAF, while black race and peri-operative administration of statins, ARBs, and ACE-Inhibitors were associated with lower risk. Patients developing POAF had significantly higher mortality along with longer and more costly hospital stays.
Our results are consistent with prior findings that black patients have a lower adjusted risk than other patients of developing atrial fibrillation generally and after cardiac surgery specifically. 23–25 This study identifies race as an important part of the risk profile for POAF in the setting of major non-cardiac surgery.
Given the large number of patients who undergo major non-cardiac surgery, even relatively infrequent peri-operative complications can lead to a substantially increased burden of morbidity and cost. As such, peri-operative management strategies that can decrease the incidence of POAF in this population have the potential to be quite impactful; in our adjusted analyses, POAF was associated with $44 million in excess costs, in addition to the attendant increase in bed-days and higher mortality. It is, however, important to note that the position of POAF in the causal chain that leads to these worsened outcomes has yet to be established.
Of note, patients with pre-existing atrial fibrillation who developed POAF had similar outcomes compared with patients who appeared to develop POAF for the first time. These findings suggest that both primary and secondary prevention strategies are worth pursuing. Whether or not they have pre-existing atrial fibrillation, special attention is warranted for patients of advanced age, those with congestive heart failure, and those undergoing high-risk procedures (i.e. thoracic or abdominal surgeries). In such patients, more active pre-operative management of volume status in CHF patients and ensuring adequate ventricular rate control in patients with chronic atrial fibrillation may be appropriately included in the pre-operative plan. While no data exist to suggest that starting these agents in the peri-operative period is beneficial, our data suggest that continuation of statins, ACE-Inhibitors and ARBs may be beneficial in suppressing arrhythmias after non-cardiac surgery.
Our study has a number of limitations. The use of the ICD-9 code 427.31 alone (rather than in combination with the code for atrial flutter: 427.32) may decrease the sensitivity of the outcome variable. However, we chose a diagnosis code of 427.31 alone to gain maximum specificity with regard to our endpoint, with the understanding that some cases of atrial fibrillation would be missed due to miscoding or misinterpretation of the ECG. Other studies of atrial fibrillation which utilized administrative data have used a similar definition of atrial fibrillation.
Although POAF was strongly associated with poor outcomes in our cohort, it is possible that POAF simply mediates the association between other postoperative illnesses (such as sepsis) and outcomes. Because we used administrative data, we may have been unable to fully control for all such confounders. In addition, our data lacked detailed information about patients’ past medical history, which prevented us from clearly discerning chronic medications from those initiated in the hospital. This also limited our ability to assess pre-operative functional status. The impact of atrial fibrillation on outcomes was attenuated by adjustment in the multi-variate regression model; given that the administrative data used for our analysis lacked clinical detail, it is possible that the inclusion of such data (were it available) would further diminish the association atrial fibrillation with poor outcomes and increased costs. We employed present-on-admission coding to discern incident from prevalent atrial fibrillation and used standard risk adjustment methodologies to account for patient risk factors. Though the use of these present-on-admission codes is becoming more common, many of the pitfalls involved with their use are related to potentially variable coding practices between hospitals. As a result of coding error, some of the cases of “new” atrial fibrillation may, in fact, have been “old” atrial fibrillation. However, these codes have recently been validated for atrial fibrillation.19 Overall, we feel that it is unlikely that the problems associated with the use of administrative data would disproportionately affect any particular patient group.
In conclusion, results from this large observational study show the incidence of POAF following major non-cardiac surgery to be about 3.0%, with considerable variation by race and surgical category. This dysrhythmia is associated with markedly increased cost and mortality, regardless of whether or not patients had pre-existing atrial fibrillation. Our data identify subsets of the study population that are at particularly high risk for POAF. Interventions aimed at preventing POAF should be prospectively tested in these higher-risk groups.
Acknowledgments
Funding: Dr. Auerbach and Ms. Maselli were supported by a Mid-Career Development grant K24HL098372 (NHLBI) during this research. Dr. Goldman was supported by an AHRQ K08 Mentored Clinical Scientist Development Award, Grant # 1 K08 HS018090-01 and NIH/NCRR UCSF-CTSI Grant Number UL1 RR024131.
Footnotes
Author potential conflicts of interest: None
Disclosures: Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. Dr. Bhave had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study. JAMA. 2001;285(18):2370–5. doi: 10.1001/jama.285.18.2370. [DOI] [PubMed] [Google Scholar]
- 2.Conen D, Chae CU, Glynn RJ, et al. Risk of death and cardiovascular events in initially healthy women with new-onset atrial fibrillation. JAMA. 2011;305(20):2080–7. doi: 10.1001/jama.2011.659. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.DeFrances CJ, Lucas CA, Buie VC, Golosinskiy A. 2006 National Hospital Discharge Survey. Natl Health Stat Report. 2008;(5):1–20. [PubMed] [Google Scholar]
- 4.Hall MJ, DeFrances CJ, Williams SN, Golosinskiy A, Schwartzman A. National Hospital Discharge Survey: 2007 summary. Natl Health Stat Report. 2010;(29):1–20. 24. [PubMed] [Google Scholar]
- 5.Mathew JP, Parks R, Savino JS, et al. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. MultiCenter Study of Perioperative Ischemia Research Group. JAMA. 1996;276(4):300–6. [PubMed] [Google Scholar]
- 6.Stanley TO, Mackensen GB, Grocott HP, et al. The impact of postoperative atrial fibrillation on neurocognitive outcome after coronary artery bypass graft surgery. Anesth Analg. 2002;94(2):290–5. doi: 10.1097/00000539-200202000-00011. table of contents. [DOI] [PubMed] [Google Scholar]
- 7.Patti G, Chello M, Candura D, et al. Randomized trial of atorvastatin for reduction of postoperative atrial fibrillation in patients undergoing cardiac surgery: results of the ARMYDA-3 (Atorvastatin for Reduction of MYocardial Dysrhythmia After cardiac surgery) study. Circulation. 2006;114(14):1455–61. doi: 10.1161/CIRCULATIONAHA.106.621763. [DOI] [PubMed] [Google Scholar]
- 8.Mitchell LB, Exner DV, Wyse DG, et al. Prophylactic Oral Amiodarone for the Prevention of Arrhythmias that Begin Early After Revascularization, Valve Replacement, or Repair: PAPABEAR: a randomized controlled trial. JAMA. 2005;294(24):3093–100. doi: 10.1001/jama.294.24.3093. [DOI] [PubMed] [Google Scholar]
- 9.Cioffi G, Cemin C, Russo TE, Pellegrini A, Terrasi F, Ferrario G. Post-discharge recurrences of new-onset atrial fibrillation following cardiac surgery: impact of low-dose amiodarone and beta-blocker prophylaxis. Ital Heart J. 2000;1(10):691–7. [PubMed] [Google Scholar]
- 10.Kapoor JR. Combination prophylactic therapy for post-operative atrial fibrillation. J Am Coll Cardiol. 2008;52(5):397. doi: 10.1016/j.jacc.2008.03.058. author reply 97–8. [DOI] [PubMed] [Google Scholar]
- 11.van Diepen S, Bakal JA, McAlister FA, Ezekowitz JA. Mortality and Readmission of Patients With Heart Failure, Atrial Fibrillation, or Coronary Artery Disease Undergoing Noncardiac Surgery: An Analysis of 38 047 Patients. Circulation. 2011;124(3):289–96. doi: 10.1161/CIRCULATIONAHA.110.011130. [DOI] [PubMed] [Google Scholar]
- 12.Polanczyk CA, Goldman L, Marcantonio ER, Orav EJ, Lee TH. Supraventricular arrhythmia in patients having noncardiac surgery: Clinical correlates and effect on length of stay. Annals of Internal Medicine. 1998;129(4):279. doi: 10.7326/0003-4819-129-4-199808150-00003. [DOI] [PubMed] [Google Scholar]
- 13.Vaporciyan AA, Correa AM, Rice DC, et al. Risk factors associated with atrial fibrillation after noncardiac thoracic surgery: analysis of 2588 patients. J Thorac Cardiovasc Surg. 2004;127(3):779–86. doi: 10.1016/j.jtcvs.2003.07.011. [DOI] [PubMed] [Google Scholar]
- 14.Bhave PD, Goldman LE, Vittinghoff E, Maselli JH, Auerbach A. Statin use and postoperative atrial fibrillation after major noncardiac surgery. Heart Rhythm. 2012;9(2):163–9. doi: 10.1016/j.hrthm.2011.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Averill R. In: All Patient Refined Diagnosis Related Groups v 20.0. Averill R, editor. 3M Corporation; 2003. [Google Scholar]
- 16.Hughes JS, Averill RF, Goldfield NI, et al. Identifying potentially preventable complications using a present on admission indicator. Health Care Financ Rev. 2006;27(3):63–82. [PMC free article] [PubMed] [Google Scholar]
- 17.Stukenborg GJ, Kilbridge KL, Wagner DP, et al. Present-at-admission diagnoses improve mortality risk adjustment and allow more accurate assessment of the relationship between volume of lung cancer operations and mortality risk. Surgery. 2005;138(3):498–507. doi: 10.1016/j.surg.2005.04.004. [DOI] [PubMed] [Google Scholar]
- 18.Stukenborg GJ, Wagner DP, Harrell FE, Jr, et al. Present-at-admission diagnoses improved mortality risk adjustment among acute myocardial infarction patients. J Clin Epidemiol. 2007;60(2):142–54. doi: 10.1016/j.jclinepi.2006.05.014. [DOI] [PubMed] [Google Scholar]
- 19.Walkey AJ, Wiener RS, Ghobrial JM, Curtis LH, Benjamin EJ. Incident stroke and mortality associated with new-onset atrial fibrillation in patients hospitalized with severe sepsis. JAMA. 2011;306(20):2248–54. doi: 10.1001/jama.2011.1615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27. doi: 10.1097/00005650-199801000-00004. [DOI] [PubMed] [Google Scholar]
- 21.Shen AY, Contreras R, Sobnosky S, et al. Racial/ethnic differences in the prevalence of atrial fibrillation among older adults--a cross-sectional study. J Natl Med Assoc. 2010;102(10):906–13. doi: 10.1016/s0027-9684(15)30709-4. [DOI] [PubMed] [Google Scholar]
- 22.Alonso A, Agarwal SK, Soliman EZ, et al. Incidence of atrial fibrillation in whites and African-Americans: the Atherosclerosis Risk in Communities (ARIC) study. Am Heart J. 2009;158(1):111–7. doi: 10.1016/j.ahj.2009.05.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rader F, Van Wagoner DR, Ellinor PT, et al. Influence of race on atrial fibrillation after cardiac surgery. Circ Arrhythm Electrophysiol. 2011;4(5):644–52. doi: 10.1161/CIRCEP.111.962670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Sun X, Hill PC, Lowery R, et al. Comparison of frequency of atrial fibrillation after coronary artery bypass grafting in African Americans versus European Americans. Am J Cardiol. 2011;108(5):669–72. doi: 10.1016/j.amjcard.2011.04.013. [DOI] [PubMed] [Google Scholar]
- 25.Lahiri MK, Fang K, Lamerato L, Khan AM, Schuger CD. Effect of race on the frequency of postoperative atrial fibrillation following coronary artery bypass grafting. Am J Cardiol. 2011;107(3):383–6. doi: 10.1016/j.amjcard.2010.09.032. [DOI] [PubMed] [Google Scholar]
- 26.Lim MJ, Roychoudhury C, Baker PL, Bossone E, Mehta RH. Differences in quality of care among patients hospitalized with atrial fibrillation as primary or secondary cause for admission. Int J Qual Health Care. 2005;17(3):255–8. doi: 10.1093/intqhc/mzi026. [DOI] [PubMed] [Google Scholar]
- 27.Jensen PN, Johnson K, Floyd J, Heckbert SR, Carnahan R, Dublin S. A systematic review of validated methods for identifying atrial fibrillation using administrative data. Pharmacoepidemiol Drug Saf. 2012;21 (Suppl 1):141–7. doi: 10.1002/pds.2317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Fry DE, Pine MB, Jordan HS, Hoaglin DC, Jones B, Meimban R. The hazards of using administrative data to measure surgical quality. Am Surg. 2006;72(11):1031–7. discussion 61–9, 133–48. [PubMed] [Google Scholar]
- 29.Glance LG, Osler TM, Mukamel DB, Dick AW. Impact of the present-on-admission indicator on hospital quality measurement: experience with the Agency for Healthcare Research and Quality (AHRQ) Inpatient Quality Indicators. Med Care. 2008;46(2):112–9. doi: 10.1097/MLR.0b013e318158aed6. [DOI] [PubMed] [Google Scholar]