The impact of 2007 infective endocarditis prophylaxis guidelines on the practice of congenital heart disease specialists

Valvular and Congenital Heart Disease

The impact of 2007 infective endocarditis prophylaxis guidelines on the practice of congenital heart disease specialists Christopher Scott Pharis, MD, a Jennifer Conway, MD, b Andrew E. Warren, MD, MSc, b Andrew Bullock, MBBS, c and Andrew S. Mackie, MD, SM a,d Alberta, and Nova Scotia, Canada; and Perth, Australia

Background The impact of the 2007 American Heart Association endocarditis prophylaxis guidelines on clinician practice has not been well established. Our objective was to evaluate how the American Heart Association endocarditis guidelines changed the practice of cardiologists who manage congenital heart disease and to ascertain the degree of practice variation among cardiologists. Methods A cross-sectional Web-based survey was e-mailed to Canadian (n = 134), Australian (n = 33), New Zealand (n = 9), and a random sample of American (n = 250) pediatric and adult congenital heart disease cardiologists in 2008. Nonrespondents received the survey 4 times by e-mail and once by regular post. Results

The response rate was 55%. The lesions for which cardiologists were most evenly divided between recommending versus not recommending prophylaxis were “rheumatic mitral stenosis of moderate severity” (45% recommended prophylaxis) and “perimembranous ventricular septal defect (VSD) status post surgical patch closure with no residual shunt 3 months post-operatively” (54% recommended prophylaxis). The lesions for which the greatest proportion of cardiologists discontinued prophylaxis were “small muscular VSD, no previous endocarditis” (80% discontinued prophylaxis) and “small audible patent ductus arteriosus” (83% discontinued prophylaxis). Only 69% recommended prophylaxis for “VSD s/p surgical patch closure with small residual shunt” despite current guidelines recommending prophylaxis for this scenario. Twenty-eight percent of respondents felt that the new guidelines leave some patients at risk, and 6% would not recounsel any low-risk patients following these guidelines.

Conclusions

The 2007 guidelines have resulted in a substantial change in endocarditis prophylaxis. There remains considerable heterogeneity among cardiologists regarding the prophylaxis of certain cardiac lesions. (Am Heart J 2011;161:123-9.)

Infective endocarditis (IE) is a life-threatening infection to which individuals with underlying cardiac pathology are predisposed.1,2 Infective endocarditis prevention has evolved over time, with the 2007 American Heart Association (AHA) guidelines3 recommending prophylaxis only for patients with cardiac conditions associated with the highest risk of adverse outcome from IE. These guidelines represent a significant departure from the 1997 AHA recommendations4 and differ from longstandFrom the aDepartment of Pediatrics, Stollery Children's Hospital and the University of Alberta, Edmonton, Alberta, Canada, bDepartment of Pediatrics, Izaak Walton Killam Health Centre and Dalhousie University, Halifax, Nova Scotia, Canada, cDepartment of Paediatrics, Princess Margaret Hospital for Children and the University of Western Australia, Perth, Australia, dDepartment of Public Health Sciences, University of Alberta, Edmonton, Alberta, Canada. Submitted February 18, 2010; accepted September 29, 2010. Reprint requests: Andrew S. Mackie, MD, SM, Stollery Children's Hospital, 4C2 Walter C. Mackenzie Centre, 8440-112th St, Edmonton, Alberta, Canada T6G 2B7. E-mail: [email protected] 0002-8703/$ - see front matter © 2011, Mosby, Inc. All rights reserved. doi:10.1016/j.ahj.2010.09.024

ing clinical practice. As such, IE prophylaxis remains a controversial topic.5-7 The impact of the 2007 AHA guidelines on the practice of pediatric and adult congenital cardiologists is unknown. A relatively high level of knowledge and acceptance of the 2007 guidelines was found among Israeli dentists, although respondents were asked about a small sample of cardiac lesions.8 Surveys before 2007 also targeted individuals without cardiology subspecialty training (patients, dentists, general practitioners).9-12 The objective of this study was to determine the influence of the 2007 AHA guidelines on recommendations made by pediatric cardiologists and adult congenital heart disease (ACHD) cardiologists regarding the prevention of IE. This study also sought to describe the variation in approaches to IE prevention within the cardiology community and between countries. We hypothesized that the 2007 guidelines had a significant impact on cardiologists' practice and that the stated practice of American cardiologists would be more conservative as compared with non-American respondents.

American Heart Journal January 2011

124 Pharis et al

Methods This study was a 4-part cross-sectional Web-based survey. Part I focused on IE prophylaxis practices in patients with a specified cardiac lesion in the context of gingival manipulation. Respondents were informed that dental procedures were performed at least 6 months after the last cardiac intervention, unless otherwise specified. The questionnaire inquired about cardiologists' recommendations both before and after the 2007 guidelines. Scenarios were chosen to represent the wide variety of cardiac lesions seen in a pediatric or ACHD cardiologist's practice and to represent both cardiac conditions for which prophylaxis before dental procedures is or is no longer recommended.3 Part II focused on IE prophylaxis practices in patients with cardiac lesions who were undergoing nondental procedures (eg, tattoo, rigid bronchoscopy). Part III focused on how the 2007 guidelines influenced counseling of patients regarding IE prevention. Part IV inquired about physician demographics. Parts I and II followed a yes/no question format, whereas Parts III and IV followed a multiple-choice or shortanswer format. The survey was piloted among 13 cardiologists at the University of Alberta (Edmonton, Alberta, Canada) and Dalhousie University (Halifax, Nova Scotia, Canada), and subsequently revised to optimize brevity and clarity, as well as face and content validity. A copy of the survey is available online at www.epicore.ualberta.ca/sbeprophylaxis.

Study population Pediatric and ACHD cardiologists in clinical practice (fulltime or part-time) were eligible for participation. Physicians identified as retired or still undergoing training were excluded. Surveys were sent to a random sample of American pediatric and ACHD cardiologists (n = 250), as well as all practicing Canadian (n = 134), Australian (n = 33), and New Zealand (n = 9) pediatric and ACHD cardiologists. American participants were randomly selected from the American Academy of Pediatrics Directory of Pediatric and Adult Congenital Cardiologists using a computerized random number generator. Canadian participants were identified through membership lists of the Canadian Adult Congenital Heart Network and the Canadian Pediatric Cardiology Association. Cardiologists in New Zealand and Australia were identified from the pediatric and congenital heart disease subgroup mailing list of the Cardiac Society of Australia and New Zealand. Study participants were contacted and informed about the survey using electronic mail. Nonresponders were e-mailed up to 4 times, with a 2-week interval between each. Continued nonresponders received a paper copy of the survey with a selfaddressed, stamped envelope. Surveys were distributed from March to July 2008 to American cardiologists, from March to June 2008 to Canadian cardiologists, and from June to December 2008 to Australian and New Zealander cardiologists. Respondents were given the option of providing their contact information and receiving a $10 coffee card honorarium (in respective countries' currencies) to reimburse them for their time. Return of a completed survey was taken as implied consent. All responses were anonymous, with no link possible between responses and e-mail addresses or other personal identifying information. This study was funded by the Stollery Children's Hospital Foundation and approved by the Health Research Ethics Board at the University of Alberta.

Sample size In the pilot phase, a decrease of ≥20% in prophylaxis rates following the 2007 AHA guidelines was observed for most clinical scenarios. To detect a 20% difference in the proportion of respondents discontinuing antibiotic prophylaxis after the 2007 guidelines with an α of .05 and a power of 99%, a total of 150 respondents was determined to be necessary. On the basis of an expected questionnaire return rate of 60% among Canadians13 and 30% among Americans,14-16 a minimum of 125 questionnaires to Canadians and 250 questionnaires to Americans was calculated to be required.

Statistical analysis Data were entered into a secure database (Microsoft SQL database, Microsoft Corporation, Redmond, WA). All data were categorical in nature. To compare cardiologist's recommendations before and after the 2007 AHA guidelines for each clinical scenario, an exact McNemar test was done. Individuals who did not provide responses to a specific question were excluded from analysis of that particular question. χ2 tests were used to determine whether country, practice type, number of years in practice, or when guidelines were last reviewed were associated with a response of not recommending IE prophylaxis for any of 6 scenarios that the 2007 guidelines recommend prophylaxis. Multiple logistic regression was then performed using a stepwise selection procedure. All tests were 2-sided and evaluated at a .05 level of significance. Analyses were conducted by the Epidemiology Coordinating and Research Centre at the University of Alberta, using SAS Version 9.1.3 (Cary, NC).

Results Respondent characteristics Of the 426 physicians mailed the questionnaire, 234 (55%) responded. Response rates by country were as follows: United States 45%, Canada 70%, Australia 58%, and New Zealand 56%. Two respondents indicated that they did not manage patients with CHD and were excluded from analyses. Two respondents completed no questions and were also excluded from analyses. Of the remaining 230 respondents, 112 were from the United States (49% of total), 94 from Canada (41%), and 24 from Australia and New Zealand (10%). Of the questionnaires analyzed, 182 (79%) were filled completely, 35 (15%) were missing 1 response, and 13 (6%) were missing ≥2 responses. Table I describes the characteristics of respondents. IE prophylaxis of dental procedures Table II describes the IE prophylaxis practices of cardiologists before and after publication of the 2007 guidelines for a variety of cardiac lesions. The scenarios for which the proportion of cardiologists recommending prophylaxis were closest to 50% were “rheumatic mitral stenosis of moderate severity” (45% of respondents recommended prophylaxis) and “perimembranous ventricular septal defect (VSD) s/p surgical patch closure with no residual shunt 3 months post-operatively” (54%

American Heart Journal Volume 161, Number 1

Pharis et al 125

Table I. Respondent characteristics

Characteristics Respondents Patient population seen Children with CHD Adults with CHD Both children and adults with CHD Duration of practice (y) b6 6-10 11-15 N15 Practice type Full-time university affiliation Private practice

Total

United States

n (%)

n (%)

n (%)

n (%)

230

112

94

24

85 (37) 27 (12) 118 (51)

22 (20)

50 (53)

13 (54)

6 (5)

20 (21)

1 (4)

84 (75)

24 (26)

10 (42)

24 (10) 38 (17) 34 (15) 134 (58)

3 (3)

16 (17)

5 (21)

16 (14)

20 (21)

2 (8)

15 (13)

15 (16)

4 (17)

78 (70)

43 (46)

13 (54)

77 (69)

84 (89)

14 (58)

25 (22)

8 (9)

2 (8)

9 (8) 1 (1)

2 (2) 0 (0)

5 (21) 3 (13)

IE prophylaxis of nondental procedures Table III reports the prophylaxis practices regarding nondental procedures.

10 (9)

5 (5)

3 (13)

9 (8) 87 (78)

7 (7) 80 (85)

0 (0) 21 (88)

6 (5)

2 (2)

0 (0)

Management of inaudible PDA and small muscular VSD Table IV describes the practice of respondents regarding the management of (a) a small inaudible PDA and (b) a small muscular VSD with a Doppler gradient of 90 mm Hg. The proportion of cardiologists that discharged a patient with a small muscular VSD from cardiac care increased following the 2007 guidelines from 11% to 28% (P b .0001). Following the 2007 guidelines, Canadian respondents were more likely to discharge these patients from cardiac care (43%) compared with respondents from Australia and New Zealand (33%) and the United States (14%) (P b .0001). The proportion that discharged a patient with an inaudible PDA from cardiac care increased from 22% to 40%, whereas the proportion that arranged for closure by cardiac catheterization decreased from 19% to 11% (P b .0001). No respondents from Australia or New Zealand stated that they would arrange for closure of inaudible PDAs, either before or after the 2007 guidelines; in comparison, following the 2007 guidelines, 7% of Canadian respondents and 17% of American respondents stated that they would arrange for closure by cardiac catheterization (P = .016).

175 (76) 35 (15) 16 (7) 4 (2)

Other⁎ Unreported Time since respondent last reviewed guidelines Before responding 18 (8) to questionnaire b2 wk ago 16 (7) ≥2 wk ago 188 (82) Unreported 8 (3)

Australia/ Canada New Zealand

septal defect (PMVSD) s/p surgical patch closure with small residual shunt” prophylaxis decreased from 100% to 69% (P b .0001); that for a “PMVSD s/p surgical patch closure with no residual shunt, 3 months post-operatively” prophylaxis decreased from 75% to 54% (P b .0001); and that for “s/p heart transplant with moderate mitral regurgitation” prophylaxis decreased from 97% to 76% (P b .0001). For “small muscular VSD, previous endocarditis,” 94% of respondents still recommended prophylaxis; and for “s/p mechanical mitral valve replacement,” 98% still recommended prophylaxis after the 2007 guidelines. Analysis of the relationship between respondent demographics (univariate χ2 test) and the answers to these scenarios showed that having read the 2007 guidelines N2 weeks before completing the survey was associated with discontinuing IE prophylaxis for ≥1 of these cardiac lesions (P = .005). Being in practice for N10 years was associated with a trend toward discontinuing prophylaxis for ≥1 of these lesions (P = .07), as was being in full-time university practice (P = .07). There was no association with the respondent's country of practice. Using stepwise logistic regression, having read the AHA guidelines N2 weeks before completing the survey was associated with discontinuing IE prophylaxis for ≥1 of these lesions (adjusted odds ratio 4.1, 95% CI 1.5-11.0, P = .005).

⁎ Eleven respondents with private practices and part-time university positions, 2 respondents with health maintenance organizations, 2 respondents with multispecialty clinics, and 1 respondent in a military medical facility.

of respondents recommended prophylaxis). The lesions for which the greatest proportion of respondents discontinued prophylaxis were “small audible patent ductus arteriosus (PDA)” (83% discontinued prophylaxis), “small muscular VSD, no previous endocarditis” (80% discontinued prophylaxis), and “valvar aortic stenosis, no previous intervention” (76% discontinued prophylaxis). We included 6 scenarios (indicated by a ⁎ in Table II) that, by our interpretation of the 2007 guidelines, continue to warrant antibiotic prophylaxis before dental procedures. The proportion of respondents recommending antibiotic prophylaxis for 4 of these lesions decreased following the 2007 guidelines. The proportion who recommended prophylaxis for “Tetralogy of Fallot (TOF) with branch pulmonary artery stenosis and stent, with moderate residual gradient” decreased from 96% to 61% (P b .0001); that for a “perimembranous ventricular

IE prophylaxis counseling practices Table V describes the change in IE prophylaxis counseling following the 2007 guidelines. For patients

American Heart Journal January 2011

126 Pharis et al

Table II. Use of IE prophylaxis for dental procedures before versus after the 2007 AHA guidelines Total

Clinical scenario PMVSD s/p surgical patch closure with small residual shunt⁎ (n = 227) PMVSD s/p surgical patch closure with no residual shunt 3 m postoperatively⁎ (n = 226) Small muscular VSD, no previous endocarditis (n = 225) Small muscular VSD, previous endocarditis⁎ (n = 224) TOF, s/p transannular patch with free pulmonary regurgitation (n = 227) TOF, s/p aortic or pulmonary homograft, without obstruction (n = 225) TOF with branch pulmonary artery stenosis and stent, with moderate residual gradient⁎ (n = 221) S/p heart transplant with moderate mitral regurgitation⁎ (n = 222) S/p mechanical mitral valve placement⁎ (n = 225) Rheumatic mitral stenosis of moderate severity (n = 224) Small audible PDA (n = 225) S/p Fontan, no valvar regurgitation, no outflow obstruction (n = 227) Valvar aortic stenosis, no previous intervention (n = 226)

United States

Before

After

Before

After

n (%)

n (%)

n (%)

n (%)

Australia/ New Zealand

Canada

P

Before

After

n (%)

n (%)

P

Before

After

n (%)

n (%)

P

227 (100) 157 (69) 110 (100)

77 (70) b.0001 94 (100) 66 (70) b.0001 23 (100) 14 (61)

.0039

169 (75)

81 (74)

58 (53) b.0001 68 (73)

51 (55) b.0001 20 (87)

13 (57)

.016

27 (12) 104 (96)

16 (15) b.0001 85 (90)

10 (11) b.0001 18 (78)

1 (4)

b.0001

207 (92) 220 (98)

122 (54)

210 (94) 107 (100) 102 (95)

.063

90 (96)

85 (90)

.125

23 (100) 23 (100)

NE

214 (94)

86 (38) 105 (95)

50 (46) b.0001 86 (92)

27 (29) b.0001 23 (100)

9 (39)

b.0001

214 (95)

144 (64) 103 (95)

62 (57) b.0001 90 (96)

70 (75) b.0001 21 (91)

12 (52)

b.0001

212 (96)

134 (61) 102 (96)

67 (63) b.0001 87 (95)

51 (55) b.0001 23 (100) 16 (70)

.016

215 (97)

168 (76) 101 (95)

86 (81) b.0001 91 (98)

63 (68) b.0001 23 (100) 19 (83)

.125

224 (100) 221 (98) 107 (99)

105 (97)

.5

94 (100) 93 (99) 1.0

23 (100) 23 (100)

NE

222 (99)

101 (45) 109 (100)

61 (56) b.0001 90 (98)

31 (34) b.0001 23 (100)

9 (39)

b.0001

219 (97) 178 (78)

33 (15) 106 (97) 92 (41) 88 (80)

24 (22) b.0001 90 (97) 43 (39) b.0001 74 (79)

8 (9) b.0001 23 (100) 43 (46) b.0001 16 (70)

1 (4) 6 (26)

b.0001 .002

225 (100)

55 (24) 109 (100)

36 (33) b.0001 93 (99)

15 (16) b.0001 23 (100)

4 (17)

b.0001

Denominators for calculating percentages were based on the nonmissing responders and varied from scenario to scenario. S/p, Status post; NE, not estimable. ⁎ Lesions requiring IE prophylaxis according to the 2007 AHA guidelines.

who no longer met the 2007 guidelines for antibiotic prophylaxis, respondents from Canada (99%) were more likely to recounsel some or all of their patients compared with respondents from the United States, Australia, or New Zealand (P = .027). A higher proportion of respondents from the United States (80%) had patients who had shown reluctance to discontinue prophylaxis as compared with Canada (66%) and Australia and New Zealand (48%) (P = .003). Respondents who allowed dentists to make recommendations about prophylaxis were more likely to have been in practice b10 years (P = .029) and were more likely to live in Australia or New Zealand (P = .008). Respondents recommending prophylaxis for every dental appointment including cleanings were more likely to be in practice N10 years (P = .024) and practice in the United States (P b .001). Practice type (full-time university vs other) had no influence on the

type of dental visits prompting recommendations for IE prophylaxis.

Discussion This is the first study to describe the impact of the 2007 AHA endocarditis prophylaxis recommendations on the practice of pediatric and ACHD cardiologists. These guidelines led to a substantial change in the practice of this physician population. This was demonstrated by a statistically significant decrease in IE prophylaxis recommended for almost every clinical scenario proposed to cardiologists from 4 countries. These guidelines also resulted in more cardiologists discharging patients with minor cardiac lesions (inaudible PDA and small muscular VSD). Despite these changes in clinical practice, more than half of respondents stated that they recounsel only

American Heart Journal Volume 161, Number 1

Pharis et al 127

Table III. Use of IE prophylaxis for nondental procedures before versus after the 2007 AHA guidelines Total

Clinical scenario S/p Fontan receiving a tattoo (n = 223) S/p Fontan undergoing rigid bronchoscopy (n = 225) Pregnant woman with a small PMVSD, no history of endocarditis, before vaginal delivery (n = 226) Patient with a PMVSD, s/p surgical patch repair and a small residual leak, requiring colonoscopy for IBD (no active GI infection) (n = 223) Patient with a mechanical mitral valve requiring cystoscopy (n = 225)

United States

Australia/ New Zealand

Canada

Before

After

Before

After

n (%)

n (%)

n (%)

n (%)

102 (46) 171 (76)

52 (23) 76 (34)

56 (53) 26 (25) b.0001 89 (82) 43 (40) b.0001

39 (42) 22 (23) b.0001 69 (73) 28 (30) b.0001

7 (30) 13 (57)

168 (74) 106 (47)

79 (72) 48 (44) b.0001

70 (75) 46 (50) b.0001

19 (83) 12 (52) .016

203 (91) 112 (50) 102 (94) 55 (51) b.0001

83 (89) 47 (51) b.0001

18 (82) 10 (45) .008

98 (90) 73 (67) b.0001

86 (92) 67 (71) b.0001

20 (91) 16 (73) .125

204 (91) 156 (69)

P

Before

After

n (%)

n (%)

P

Before

After

n (%)

n (%)

P

4 (17) .25 5 (22) .008

Denominators for calculating percentages were based on the nonmissing responders and varied from scenario to scenario. IBD, Inflammatory bowel disease; GI, gastrointestinal.

Table IV. Clinical practice of respondents regarding the management of small inaudible PDA and small muscular VSD before versus after the 2007 AHA guidelines

Clinical practice Small muscular VSD Discharged from cardiac care Small inaudible PDA Discharged from cardiac care Arranged cardiac catheterization closure Arranged cardiac follow-up

Total (N = 228)

United States (n = 110)

Before

After

Before

After

n (%)

n (%)

n (%)

n (%)

25 (11)

63 (28)

4 (4)

51 (22) 43 (19) 134 (59)

92 (40) 26 (11) 110 (48)

20 (18) 29 (26) 61 (55)

some or none of their patients in response to the guidelines, and more than one quarter of respondents felt that the 2007 guidelines leave some patients at risk for acquiring IE. We observed a significant decrease in rates of prophylaxis for 4 of 6 scenarios that by our group's interpretation of the 2007 guidelines still warrant IE prophylaxis. Although the recommendation for IE prophylaxis for patients with a prosthetic cardiac valve and for those with previous IE remains widely practiced, prophylaxis of other scenarios is controversial. This is highlighted by the scenario “PMVSD s/p surgical patch closure with no residual shunt, 3 months post-operatively” (Table II). Only 54% of cardiologists indicated that they would recommend prophylaxis to these individuals. However, the AHA guidelines state that patients with “completely repaired CHD with prosthetic material … during the first 6 months after the procedure” should continue to receive IE

Canada (n = 94) Before

After

P

n (%)

n (%)

15 (14)

.001

18 (19)

39 (35) 19 (17) 52 (47)

b.0001 .002 .061

23 (25) 14 (15) 57 (61)

Australia/ New Zealand (n = 24) Before

After

P

n (%)

n (%)

P

40 (43)

b.0001

3 (13)

8 (33)

.025

42 (45) 7 (7) 45 (48)

b.0001 .008 .003

8 (33) 0 (0) 16 (67)

11 (46) 0 (0) 13 (54)

.083 NE .083

prophylaxis.3 Whether physician noncompliance is due to unclear recommendations, differing interpretation of the literature, or other reasons is uncertain. Cardiologists who had not read the 2007 AHA guidelines for at least 2 weeks before responding to the survey were more likely to discontinue prophylaxis for scenarios in which prophylaxis is still recommended. No demographic features were independently predictive of a physician group that was noncompliant with the AHA recommendations. This finding implies that physicians have difficulty recalling the details of the guidelines and suggests that it may be beneficial to post the AHA recommendations in patient areas (eg, ambulatory clinics) to quickly and easily refer to the recommendations when needed. There are several lesions for which a substantial proportion of cardiologists continue to prescribe IE prophylaxis despite the 2007 guidelines recommending

American Heart Journal January 2011

128 Pharis et al

Table V. Respondents IE prophylaxis counseling practices following the 2007 AHA guidelines Total (N = 228) United States (n = 111) Canada (n = 94) Counseling practices Counseling of patients no longer meeting 2007 criteria for IE prophylaxis Recounseled all patients Recounseled some patients Did not recounsel patients Reasons for not recounseling patients⁎ Belief that new recommendations leave some patients at risk Fear of litigation Lack of awareness that new guidelines existed Violation of long-standing practice pattern and/or patient expectations Forgot to recounsel patients Other† Care for patients who have shown reluctance to stop IE prophylaxis Patient reasons for being reluctant to discontinue prophylaxis⁎ No previous allergic reaction to endocarditis prophylaxis medication Patient or parent anxiety about endocarditis Other‡ Prophylaxis recommendations and dental procedure invasiveness⁎ Allow the dentist to make recommendations based on the planned dental procedure Recommend prophylaxis for all dental visits, including “cleaning” Prophylaxis only if the dentist anticipates gingival bleeding Provide other recommendations§

Australia/New Zealand (n = 23)

n (%)

n (%)

n (%)

n (%)

111 (49) 104 (46) 13 (6)

47 (42) 55 (50) 9 (8)

54 (57) 39 (42) 1 (1)

10 (44) 10 (44) 3 (13)

63 (28)

35 (32)

24 (26)

4 (17)

4 (2) 1 (0) 33 (15)

3 (3) 1 (1) 21 (19)

0 (0) 0 (0) 9 (10)

1 (4) 0 (0) 3 (13)

11 (5) 10 (4) 162 (71)

3 (3) 4 (4) 89 (80)

6 (6) 3 (3) 62 (66)

2 (9) 3 (13) 11 (48)

44 (19)

26 (23)

17 (18)

1 (4)

152 (67) 11 (5)

84 (76) 4 (4)

58 (62) 6 (6)

10 (44) 1 (4)

85 (38)

30 (28)

42 (46)

13 (54)

70 (32)

46 (43)

23 (25)

1 (4)

75 (34)

33 (31)

31 (34)

11 (46)

19 (9)

10 (9)

9 (10)

0 (0)

⁎ Respondents were allowed to select N1 option. † Believe the guidelines require physician clinical judgment; have not had the opportunity to recounsel all patients; allow patients to decide; do not recounsel to avoid patient confusion; no previous adverse reaction to antibiotic; awaiting Australian guidelines. ‡ Poor evidence to discontinue prophylaxis; safer to continue prophylaxis; dentist/anesthetist requests prophylaxis; always used prophylaxis so why stop; uncertain of new guidelines. § Highly variable individual responses that can be provided upon request.

that doing so is not necessary. For example, 50% of respondents indicated that they would recommend prophylaxis for a patient undergoing colonoscopy, and 69% reported that they would recommend prophylaxis for a patient with a mechanical mitral valve undergoing cystoscopy. These responses may reflect the influence of 2006 British guidelines recommending prophylaxis for some gastrointestinal and genitourinary procedures.17 Our study has demonstrated that the AHA guidelines have had an impact on the practice of cardiologists internationally. Not only American respondents but also respondents from Canada, Australia, and New Zealand changed their clinical practice, with some variation in the management of lesions such as small muscular VSDs and inaudible PDAs. This international change in practice is in keeping with recent literature from Canada and Australia that has largely been supportive of and consistent with the AHA recommendations.18-20

Our study has several limitations. The survey was designed for the purpose of this study and has not been previously validated. We attempted to optimize reliability and validity with local pilot testing of the instrument. Because of repeated statistical testing, significant P values may be due to chance alone. However, the magnitude of the before/after differences suggests that significant P values are real rather than due to a type I error. Given the impracticality of a chart review from physician's offices, we were unable to confirm that respondents actually do what they report. We could not determine the characteristics of nonresponders, and a responder bias may exist. The survey was distributed b1 year after publication of the revised IE guidelines, which may have been insufficient time for some cardiologists to change clinical practice. Physicians' attitudes toward the 2007 guidelines were not thoroughly explored.

American Heart Journal Volume 161, Number 1

Conclusion The 2007 AHA guidelines on the prevention of IE have resulted in a substantial change in the endocarditis prophylaxis practices of pediatric and ACHD cardiologists. These changes have resulted in a decrease in IE prophylaxis for most cardiac lesions. However, a substantial proportion of physicians and their patients have expressed reluctance at adopting the 2007 guidelines. As a consequence, there remains considerable heterogeneity in the cardiology community regarding the prophylaxis of certain cardiac lesions. Further studies are required to elucidate the reasons for this variability in clinical practice.

References 1. Morris CD, Reller MD, Menashe VD. Thirty-year incidence of infective endocarditis after surgery for congenital heart defect. JAMA 1998; 279:599-603. 2. Li W, Somerville J. Infective endocarditis in the grown-up congenital heart (GUCH) population. Eur Heart J 1998;19:166-73. 3. Wilson W, Taubert KA, Gewitz M, et al. Prevention of infective endocarditis: guidelines from the American Heart Association. Circulation 2007;116:1736-54. 4. Dajani AS, Taubert KA, Wilson W, et al. Prevention of bacterial endocarditis. Recommendations by the American Heart Association. JAMA 1997;277:1794-801. 5. Cheng TO. Endocarditis prophylaxis in patients with mitral valve prolapse remains a controversial issue despite the new American Heart Association guidelines. Int J Cardiol 2008;127:149-50. 6. Prendergast BD, Harrison JL, Naber CK. Commentary on endocarditis prophylaxis: a quaint custom or medical necessity? Heart 2008;94: 931-4. 7. Shanson D. New British and American guidelines for the antibiotic prophylaxis of infective endocarditis: do the changes make sense? A critical review. Curr Opin Infect Dis 2008;21:191-9.

Pharis et al 129

8. Zadik Y, Findler M, Livne S, et al. Dentists' knowledge and implementation of the 2007 American Heart Association guidelines for prevention of infective endocarditis. 2008;106:e16-9. 9. Lauber C, Lalh SS, Grace M, et al. Antibiotic prophylaxis practices in dentistry: a survey of dentists and physicians. J Can Dent Assoc 2007; 73:263e. 10. Niwa K, Nakazawa M, Miyatake K, et al. Survey of prophylaxis and management of infective endocarditis in patients with congenital heart disease: Japanese nationwide survey. 2003;67:585-91. 11. Seto TB, Thomas L, Baden LR, et al. Specialty and training differences in the reported use of endocarditis prophylaxis at an academic medical center. Am J Med 2001;111:657-60. 12. Seto TB, Kwiat D, Taira DA, et al. Physicians' recommendations to patients for use of antibiotic prophylaxis to prevent endocarditis. JAMA 2000;284:68-71. 13. Conway J, Wong KK, O'Connell C, et al. Cardiovascular risk screening before starting stimulant medications and prescribing practices of Canadian physicians: impact of the Health Canada advisory. Pediatrics 2008;122:e828-34. 14. Kahwaji IY, Connuck DM, Tafari N, et al. A national survey on the pediatric cardiologist's clinical approach for patients with Kawasaki disease. Pediatr Cardiol 2002;23:639-46. 15. Khalid O, Luxenberg DM, Sable C, et al. Aortic stenosis: the spectrum of practice. Pediatr Cardiol 2006;27:661-9. 16. Wong KK, Potts JE, Etheridge SP, et al. Medications used to manage supraventricular tachycardia in the infant a North American survey. Pediatr Cardiol 2006;27:199-203. 17. Gould FK, Elliott TS, Foweraker J, et al. Guidelines for the prevention of endocarditis: report of the working party of the British Society for Antimicrobial Chemotherapy. J Antimicrob Chemother 2006;57: 1035-42. 18. Embil JM, Chan KL. The American Heart Association 2007 endocarditis prophylaxis guidelines: a compromise between science and common sense. Can J Cardiol 2008;24:673-5. 19. Moulds RF, Jeyasingham MS. Antibiotic prophylaxis against infective endocarditis: time to rethink. Med J Aust 2008;189:301-2. 20. Infective Endocarditis Prophylaxis Working Group. Prevention of endocarditis. Melbourne: Therapeutic Guidelines Limited; 2008.