Author + information
- Received December 23, 2014
- Revision received March 3, 2015
- Accepted March 9, 2015
- Published online May 19, 2015.
- Sadip Pant, MD∗,
- Nileshkumar J. Patel, MD†,
- Abhishek Deshmukh, MD‡,
- Harsh Golwala, MD∗,
- Nilay Patel, MD§,
- Apurva Badheka, MD‖,
- Glenn A. Hirsch, MD, MHS∗ and
- Jawahar L. Mehta, MD, PhD¶∗ ()
- ∗Department of Cardiovascular Medicine, University of Louisville, Louisville, Kentucky
- †Department of Internal Medicine, Staten Island University, Staten Island, New York
- ‡Department of Cardiovascular Medicine, Mayo Clinic, Rochester, Minnesota
- §Department of Internal Medicine, Saint Peter’s University Hospital, Jersey City, New Jersey
- ‖Department of Cardiovascular Medicine, Yale New Haven Medical Center, New Haven, Connecticut
- ¶Department of Cardiovascular Medicine, University of Arkansas for Medical Sciences, Little Rock, Arkansas
- ↵∗Reprint requests and correspondence:
Dr. Jawahar L. Mehta, Department of Cardiovascular Medicine, University of Arkansas for Medical Sciences, 4301 West Markham Street, Little Rock, Arkansas 72205.
Background In accordance with the 2007 American College of Cardiology and American Heart Association infective endocarditis (IE) guideline update, antibiotic prophylaxis is now being restricted to a smaller number of cardiac conditions with very high risk for adverse outcomes from IE. However, there is scant data on IE trends since this major practice change in the United States.
Objectives The aim of this study was to compare temporal trends in IE incidence, microbiology, and outcomes before and after the change in the 2007 IE prophylaxis guideline in the United States.
Methods The NIS (Nationwide Inpatient Sample) database was used to investigate IE hospitalization rates in the United States from 2000 through 2011. The mean annual rates of IE before and after the 2007 guideline change were compared using segmented regression analysis.
Results There were 457,052 IE-related hospitalizations in the United States from 2000 to 2011, with a steady increase in incidence (p < 0.001). The trend in IE hospitalization rates from 2000 to 2007 and from 2008 to 2011 was not significantly different (p = 0.74). The increases in the number of Staphylococcus IE cases per million population during the study periods 2000 to 2007 and 2008 to 2011 were similar (p = 0.13), but Streptococcus IE hospitalization rates were significantly higher after the release of new guidelines (p = 0.002). Finally, valve replacement rates for IE steadily increased from 2000 to 2007 (p = 0.03) but showed a plateau from 2007 to 2011. Overall, there was no significant difference in the rates of valve replacement for IE before and after the release of new guideline (p = 0.23).
Conclusions These results show that IE incidence has increased in the United States over the past decade. With regard to the microbiology of IE, there has been a significant rise in the incidence of Streptococcus IE since the 2007 guideline revisions. However, the rates of hospitalization and valve surgery for IE have not increased since the change in IE prophylaxis guideline in 2007.
The epidemiology of infective endocarditis (IE) has changed over the years because of changes in the prevalence of risk factors, as well as improved diagnostic tools and management. Although there has been a reduction in rheumatic heart disease in United States (1), there has been an increase in invasive procedures as well as an increase in high-risk groups, such as intravenous drug users and those with human immunodeficiency virus and diabetes mellitus (2–5). Furthermore, the increase in the survival of IE risk–prone populations, such as adults with congenital heart disease and prosthetic device implants, also contributes to increasing IE incidence. Finally, global trends in infectious disease epidemiology also may affect trends in IE. A revised guideline for antibiotic prophylaxis for the prevention of IE was released in 2007 by the American Heart Association (AHA) and the American College of Cardiology (ACC) (6). According to the new guidelines, antibiotic prophylaxis is restricted to only a small number of cardiac conditions with high risk for adverse outcomes from IE. However, there is a paucity of data on IE trends since this major change in practice guidelines in the United States.
Through our study, we sought to: 1) identify trends in the incidence of IE before and after the publication of the new guideline for antibiotic prophylaxis for IE in 2007; 2) assess trends in the microbiology of IE over the past decade; and 3) assess the trend in valve replacement before and after the change in the guideline.
We performed a retrospective observational cohort study, using the Healthcare Cost and Utilization Project NIS (Nationwide Inpatient Sample) database, sponsored by the Agency for Healthcare Research and Quality (7). We used the International Classification of Diseases, Ninth Revision, Clinical Modification code to identify patients discharged with acute and subacute bacterial endocarditis between 2000 and 2011 (Online Appendix). Similarly, the microbiology, as well as valve replacement, were identified using appropriate procedure codes. With regard to microbiology, 4 groups were included: Staphylococcus, Streptococcus, Gram negative (GN), and fungi. We excluded endocarditis due to syphilis, rheumatic heart disease (without infection), gonococcal endocarditis, and lupus or other noninfectious causes. We used weights provided with the NIS to generate national estimates.
The NIS database is the largest all-payer database (Medicare, Medicaid, private insurance, and uninsured) of hospital inpatient stays available in United States (excluding the federal, institutional, and short-term rehabilitation hospitals). Each year, the NIS data are updated to approximately represent a 20% stratified sample of U.S. hospitals. Each individual hospitalization is deidentified and maintained in the NIS as a unique entry with 1 primary discharge diagnosis and ≤24 secondary diagnoses during that hospitalization. Each entry also contains information on demographic details, including age, sex, race, insurance status, primary and secondary procedures (up to 14), hospitalization outcome, total charges, and length of stay. Data from the NIS have been previously used to identify, track, and analyze national trends in health care utilization, patterns of major procedures, trends in hospitalizations, charges, quality, and outcomes (8–10). They also have been used to study the epidemiology of IE in the United States in the past (11,12).
We used Stata IC version 11.0 (StataCorp LP, College Station, Texas) and SAS version 9.2 (SAS Institute Inc., Cary, North Carolina) for the analyses, which accounted for the complex survey design and clustering. To estimate the annual rates of IE hospitalizations, we divided the total number of IE cases in a given year by the U.S. census population for that year, which were represented in tables and graphs per 100,000 or per million population. The proportion of IE hospitalizations due to each organism was expressed in 2 ways: 1) as a proportion of all IE hospitalizations; and 2) per population for that year. We compared the estimated mean annual rates of IE for data from before and after the introduction of the 2007 ACC/AHA IE antibiotic prophylaxis guidelines using piecewise regression analysis (also known as segmented regression analysis) of the interrupted time series (13). The statistical significance level was set at p ≤ 0.05.
A total of 457,052 estimated IE hospitalizations were identified during the study period (2000 to 2011). The annual IE hospitalization rate, microbiology, and valve replacement rates in the United States from 2000 to 2011 are summarized in Table 1.
In the entire cohort, there was a steady increase in the incidence of IE hospitalizations from 2000 to 2011 (p < 0.001). The change in IE hospitalization rate for 2000 to 2007 was 0.54 per 100,000 population (95% confidence interval: 0.32 to 0.75; p < 0.001). The change in IE hospitalization rate for 2007 to 2011 was 0.6 per 100,000 population (95% confidence interval: 0.23 to 0.97; p = 0.005). The trends in IE hospitalization rates per 100,000 population between the study periods 2000 to 2007 and 2008 to 2011 were not significantly different (p = 0.74) in the United States (Figure 1).
The microbiology data demonstrated a steady rise in number of Staphylococcus IE cases. The proportion of IE due to Staphylococcus (expressed as percentage of total IE cases) increased from 33% in 2000 to 40% in 2011, a relative increase of 18.9% (p < 0.001) (Figure 2). The proportion of IE due to Streptococcus also increased from 24.8% in 2000 to 27% in 2011, a relative increase of 10.6% (p < 0.001) (Figure 2). During the study period, the proportion of IE due to GN bacteria increased from 5.3% to 8.2% (a relative increment of 33.9%; p < 0.001), and IE due to fungi increased from 0.6% to 1.4% (a relative increment of 59.6%; p < 0.001).
The increases in the number of Staphylococcus IE cases per million population during the study period 2000 to 2007 and 2008 to 2011 were similar (p = 0.13) (Central Illustration). There was, however, a statistically significant increase in Streptococcus IE cases comparing the time periods before (2000 to 2007) and after (2008 to 2011) the release of new guidelines (p = 0.002) (Central Illustration).
The valve replacement rates for IE steadily increased from 2000 to 2007 (p = 0.03), followed by a plateau from 2007 to 2011. There was no significant difference in valve replacement rates for IE in the United States after the inception of the 2007 guideline (p = 0.23) (Figure 3). The trends in valve replacement for Staphylococcus IE and Streptococcus IE did not change significantly before and after 2007 (p = 0.13 and p = 0.49, respectively) (Figure 4).
Our study has several important findings. First, there has been a steady increase in the incidence of IE hospitalizations over the past decade in the United States. However, the incidence of IE pre- and post-inception of new antibiotic prophylaxis guidelines is not significantly different (Central Illustration). In parallel to these findings, the rate of valve replacement for IE did not change after the release of new guidelines in 2007. Second, the increase in IE incidence was seen across all types of pathogens: Staphylococcus, Streptococcus, GN bacteria, and fungi. Finally, the rate of Streptococcus IE–related hospitalization increased significantly after the release of new guideline in the United States, while Staphylococcus IE hospitalizations, although also on the rise, did not differ significantly after the 2007 ACC/AHA guideline update.
The increase in the U.S. incidence of IE over the past decade has been reported previously and is likely related to an increase in the sizes of at-risk populations, such as older, diabetic, and hemodialysis patients (14,15). In addition, the number of invasive procedures leading to transient bacteremia has increased markedly over this time period (2). Finally, there has been increase in survival in IE risk–prone populations, such as adults with congenital heart disease and prosthetic device implants, leading to an increase in IE incidence. Despite the overall increase in IE hospitalizations from 2000 to 2011, we did not observe any significant change in hospitalization rates from the pre-guideline era (2000 to 2007) to the post-guideline era (2007–2011). Similar findings have been replicated in France, where a guideline change in 2002 did not result in an increase in IE cases (16). However, a recent study from the United Kingdom showed an increase in IE cases with a decrease in antibiotic use, after a National Institute for Health and Care Excellence guideline change in 2008 (17). We speculate that the difference in microbiology of IE among these populations may partly account for such differences. Staphylococcus has remained the predominant pathogen for IE in the United States. Because the antibiotic prophylaxis in the guideline does not routinely cover Staphylococcus, its growing trend since 2000 would not be affected by the change in the 2007 guideline.
Our study shows that Streptococcus IE cases increased significantly after the change in the guideline in 2007. This is in contrast to prior studies, which showed no inflection in hospitalization rates after the 2007 ACC/AHA changes in prophylaxis recommendations (11,12). The prior studies, however, had limited follow-up duration (2-year follow-up after the guideline change). A similar observation was made in the U.K. study, in which a steady incidence of IE was noted for first 2 years after publication of the guidelines (18).
The increasing incidence of IE from 2000 to 2011 was seen across all 4 pathogens associated with IE, namely, Staphylococcus, Streptococcus, GN bacteria, and fungi. Specifically, the number of staphylococcal IE cases increased, with a 20% relative rise during the 11-year study period. Staphylococcus has been found to be a predominant etiology for health care–associated native valve IE in the United States for nosocomial as well as nonnosocomial acquisition (2). Increased contact with health care providers, an increase in invasive procedures associated with bacteremia, and increases in the sizes of high-risk population, such as the elderly and patients with diabetes mellitus or end-stage renal disease, may well be responsible for growing rates of staphylococcal IE cases (2,3,5,19). In addition, there has been an increasing trend in the use of prosthetic cardiac devices (pacemakers, defibrillators, and prosthetic cardiac valves) in the United States (20), which make patients susceptible to staphylococcal infections. However, the rates of rise in Staphylococcus IE before and after the guideline change did not differ significantly.
We also found that the overall rates of valve replacement for IE in the United States have stabilized since 2007. This was observed for both Streptococcus and Staphylococcus IE. This may reflect the increasing awareness of this disease and better diagnostic tools, leading to earlier detection of IE in the United States. As a result, fewer patients may be developing severe complications from this disease that require surgical intervention. This is supported by our study, in which we noticed that rates of valve replacement for Staphylococcus IE changed significantly before 2007 (p = 0.05). However, the trend stabilized after 2007 (p = 0.78) (Figure 3).
Despite its national scope and the use of a widely used, well-characterized database, our study had some important limitations. The main one is the lack of proper validation studies for the International Classification of Diseases codes for IE. However, there has been a pilot study with a broader extraction of these codes for IE, which had a positive predictive value of 81% (21). The trend data also are affected by variations in coding practices among hospitals. Hence, it is not possible to know whether the increase in the incidence of Streptococcus IE is due to more complete recording or a true increase. However, there was no unexpected increase seen in IE due to Staphylococcus, GN bacteria, or fungi, which at least points toward the latter speculation that this observation might be something more than just a coding issue. Furthermore, the NIS, being a discharge-level database, cannot distinguish if multiple hospitalizations are from the same patient. This may limit the precision of our estimates. Finally, our database does not capture antibiotic use, and hence, we do not know the impact of the 2007 guideline change on antibiotic prescription for IE prophylaxis. Despite these limitations, the NIS database provides data from approximately 20% of all U.S. community hospitals. This large cohort, representative sample, and longer follow-up duration after the 2007 guideline change makes our study reflective of changing trends in IE hospitalizations from a national perspective.
The overall incidence of IE showed a steady increase in the United States from 2007 to 2011. However, the annual hospitalization rates and frequency of valve surgery due to IE have not changed significantly since the change in IE prophylaxis guideline in 2007. With regard to the microbiology of IE, there has been a significant rise in the incidence of Streptococcus IE since the 2007 guideline revisions. We speculate that this could be related to the decrease in the use of IE antibiotic prophylaxis since the guideline change. A prospective study designed to understand the antibiotic prescription pattern after the change in IE guidelines is needed to further test the validity of our observations. Furthermore, ongoing monitoring of the impact of these guideline recommendations is highly warranted, as the new guideline in itself is not “evidence based” but rather attempts to demonstrate the lack of evidence underlying the previous recommendations for IE prophylactic antibiotics.
COMPETENCY IN MEDICAL KNOWLEDGE: Unnecessary use of antibiotic drugs exposes patients to unnecessary side effects and leads to the development antibiotic-resistant bacteria.
COMPETENCY IN PATIENT CARE: Since the publication of the 2007 ACC/AHA scientific statement on IE, antibiotic prophylaxis is recommended for a more restricted number of cardiac conditions in which IE is associated with a high risk for adverse outcomes.
TRANSLATIONAL OUTLOOK: More research is needed to define the conditions for which antibiotic prophylaxis is actually effective in preventing IE and to assess the impact of changes in the practice of antibiotic prophylaxis on the epidemiology of IE.
For a list of the ICD codes identifying infective endocarditis and causative organisms, please see the online version of this article.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose. All authors contributed equally to this work.
- Abbreviations and Acronyms
- American College of Cardiology
- American Heart Association
- Gram negative
- infective endocarditis
- Received December 23, 2014.
- Revision received March 3, 2015.
- Accepted March 9, 2015.
- 2015 American College of Cardiology Foundation
- Wilson W.,
- Taubert K.A.,
- Gewitz M.,
- et al.
- ↵Agency for Healthcare Research and Quality. Introduction to the HCUP Nationwide Inpatient Sample. Available at: http://www.hcup-us.ahrq.gov/db/nation/nis/NIS_2007_INTRODUCTION.pdf. Accessed November 12, 2013.
- Desimone D.C.,
- Tleyjeh I.M.,
- Correa de Sa D.D.,
- et al.
- Boyle J.P.,
- Honeycutt A.A.,
- Narayan K.M.,
- et al.
- Wiener J.M.,
- Tilly J.
- Duval X.,
- Delahaye F.,
- Alla F.,
- et al.
- Thornhill M.H.,
- Dayer M.J.,
- Forde J.M.,
- et al.