You are here

Article Text

Article menu
Valvular heart disease
Original article
Early versus late surgical intervention or medical management for infective endocarditis: a systematic review and meta-analysis
  1. Mahesh Anantha Narayanan1,
  2. Toufik Mahfood Haddad1,
  3. Andre C Kalil2,
  4. Arun Kanmanthareddy3,
  5. Rakesh M Suri4,
  6. George Mansour1,
  7. Christopher J Destache5,
  8. Janani Baskaran6,
  9. Aryan N Mooss3,
  10. Tammy Wichman7,
  11. Lee Morrow7,
  12. Renuga Vivekanandan8
  1. 1Department of Internal Medicine, Creighton University School of Medicine, Omaha, Nebraska, USA
  2. 2Division of Infectious Diseases, University of Nebraska School of Medicine, Omaha, Nebraska, USA
  3. 3Division of Cardiology, Creighton University School of Medicine, Omaha, Nebraska, USA
  4. 4Department of Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cleveland, Ohio, USA
  5. 5School of Pharmacy & Health Professions and School of Medicine, Creighton University, Omaha, Nebraska, USA
  6. 6University of Texas Southwestern at Dallas, Dallas, Texas, USA
  7. 7Division of Pulmonary Critical Care and Sleep Medicine, Creighton University School of Medicine, Omaha, Nebraska, USA
  8. 8Division of Infectious Diseases, Creighton University School of Medicine, Omaha, Nebraska, USA
  1. Correspondence to Dr Mahesh Anantha Narayanan, Department of Internal Medicine, Creighton University School of Medicine, Omaha, NE 68154, USA; mahesh_maidsh{at}yahoo.com

https://doi.org/10.1136/heartjnl-2015-308589

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Video abstract

    Introduction

    Infective endocarditis (IE) remains a major medical illness with a high mortality approaching 50%1 secondary to complications including congestive heart failure and neurological events.2 The American College of Cardiology/American Heart Association (ACC/AHA) endocarditis guidelines and the European Society of Cardiology (ESC) Endocarditis guidelines recommend specific parameters to be met for performing early valve surgery.3 ,4 Indeed, recommending early surgery is a class-IIa indication for patients with recurrent emboli and persistent vegetation;3 the European guidelines recommend early surgery as a class-IIb indication for patients with >1.5 cm vegetation.4 The decision to perform early surgery has always been a challenge, considering the associated complications and unpredictable response to antibiotic therapy. There are only a limited number of randomised trials to compare the effect of early surgery with conventional therapy.5 We performed a systematic review and meta-analysis with available evidence to determine the benefits, risks and the optimal timing of early surgery in IE.

    Materials and methods

    Literature search

    We performed an electronic search using the terms ‘early surgery’, ‘valve surgery’, ‘endocarditis’, ‘bacterial endocarditis’ and ‘infective endocarditis’ using PubMed, Cochrane, EMBASE, CINAHL and Google-scholar databases for studies published between January 1960 and April 2015, comparing early valve surgery with conventional therapy for IE. Studies included ranged from January 1996 to January 2015. The detailed search strategy for PubMed is shown in online supplementary appendix 1.

    The systematic review and meta-analysis was performed according to the MOOSE (Research-Checklist) guidelines.6 Search strategy is shown as PRISMA flowchart7 (figure 1). We also reviewed the reference sections of all included studies, review articles and editorials for completeness. According to Cochrane guidelines, we excluded conference abstracts. To evaluate the quality of studies, we used the Newcastle–Ottawa scale8 (see online supplementary table S1). We defined best-quality studies with a score of 9 (maximum), and the highest-quality studies by a score of 8.

    Figure 1
    Figure 1

    PRISMA flowchart.

    Eligibility criteria

    Studies included should (1) be a randomised controlled trial, retrospective or prospective observational study; (2) compared two groups, one with early valve surgery either at 20 days or less or during initial hospitalisation, and the other with medical therapy with or without late surgery; (3) included only adult patients (excluded patients <16 years); (4) be published in peer-reviewed literature and (5) be in English language.

    Definitions

    All-cause mortality was defined as death due to any cause at follow-up. Early surgery was defined as surgery at 20 days or less of diagnosis of IE or during initial hospitalisation. Conventional therapy was defined as either medical therapy or late surgery (at >20 days). We chose 20 days as cut-off since most studies comparing early surgery with conventional therapy had surgery performed within 3 weeks of diagnosis (see online supplementary table S1). In the subgroup analysis, we compared all-cause mortality for (1) early surgery at 7 days or less with conventional therapy and (2) early surgery between 8 and 20 days with conventional therapy. In the second subgroup analysis, we compared all-cause mortality for surgery within a 20-day time period with (1) surgery at greater than 20 days and to (2) medical therapy.

    Data collection

    We included patient demographics, sample size and type of study in a structured abstract. Using a structured form, we abstracted patient characteristics including setting, sample size, proportion of cases undergoing valve surgery, mortality data and risk estimates for different analyses. Two reviewers (MAN and TMH) independently collected and abstracted the data after reviewing full-text articles. Any initial disagreement was reviewed and resolved by consensus.

    Outcome

    The study's primary end point was all-cause mortality. Secondary end points included in-hospital death or 30-day mortality, embolic events, heart failure and recurrence of endocarditis.

    Statistical analysis

    Statistical analysis was performed using Comprehensive Meta-analysis (CMA) (V.3.3.070). Categorical data were presented as odds ratio (OR) with 95% CIs using the random-effects model for outcomes with different effect sizes. Early surgery was considered as the experimental group and hence any OR <1 favours early surgery. Publication bias was analysed visually with a funnel plot. Cochrane's Q statistics were calculated and used to determine the heterogeneity of included studies. I2 values of 25%, 50% and 75% were considered as low, moderate and high heterogeneity, respectively.9 A sensitivity analysis was included when necessary. A p value of <0.05 was considered significant.

    Results

    Study characteristics

    Two independent reviewers (MAN and TMH) performed the initial search (figure 1); and 1428 unique studies were identified (see online supplementary appendix 1). Abstracts of these studies were screened for eligibility and included in the analysis after mutual consensus of the two reviewers. Finally, 21 studies5 ,10–29 that met the eligibility criteria were included. A total of 4797 patients underwent early surgery at 20 days or less and 6251 patients received conventional therapy. Patient demographic data and baseline characteristics are shown in on-line supplementary tables S1 and S2 respectively. Pooled OR of all outcomes is summarised in table 1.

    View this table:
    Table 1

    Summary of overall results from the analysis

    All-cause mortality of all included studies

    All-cause mortality was mentioned in 21 studies. In the group that had early surgery, all-cause mortality was significantly lower than in the group without early surgical intervention (OR 0.61, 95% CI 0.50 to 0.74, p<0.001) (figure 2). Heterogeneity was high among the included studies (I2=68%). Publication bias assessed by the funnel plot showed minimal bias, with most studies centred symmetrically at the top (p=0.80) (see online supplementary figure S1). A sensitivity analysis excluding the study27 with maximum weight did not alter the results (OR 0.58, 95% CI 0.49 to 0.69, p<0.001). Another sensitivity analysis done by excluding studies with a low Newcastle–Ottawa scale (7 or less) still showed odds favouring early surgical group (OR 0.64, 95% CI 0.51 to 0.79, p<0.001).

    Figure 2
    Figure 2

    Comparison of all-cause mortality between early surgery (at 20 days or less) and conventional therapy (late surgery at >20 days or medical therapy).

    Second, we categorised studies wherein surgery was performed (1) at 7 days or less and (3) between 8 and 20 days and compared these groups with conventional therapy (figure 3A, B). We chose 7-day time period as cut-off since previous data showed higher risk of embolic events in the first week after diagnosis of IE and evidence showing failure of medical treatment for persistent bacteraemia after 1 week.3 ,30 Most studies included in the subanalysis were distributed around this number. All-cause mortality in the group where surgery was performed at 7 days or less was estimated to be significantly lower (OR 0.61, 95% CI 0.39 to 0.96, p=0.034) when compared with conventional therapy. In the subgroup where surgery was performed at 8–20 days, odds of survival were still in favour of early surgery (OR 0.64, 95% CI 0.48 to 0.86, p=0.003) when compared with conventional therapy.

    Figure 3
    Figure 3

    (A) Comparison of all-cause mortality between early surgery at 7 days or less and conventional therapy; (B) comparison of all-cause mortality between early surgery at 8–20 days and conventional therapy; (C) comparison of all-cause mortality between early surgery (at 20 days or less) and late surgery (at >20 days); (D) comparison of all-cause mortality between early surgery (at 20 days or less) and medical therapy.

    Third, in the subanalysis comparing early surgery (at 20 days or less) with (1) late surgery (at >20 days) and (2) medical therapy (figure 3C, D), all-cause mortality was lower with early surgery when compared with late surgery (>20 days) but without achieving a statistical significance (OR 0.87, 95% CI 0.43 to 1.73, p=0.685). All-cause mortality was lower in the early surgical group (OR 0.58, 95% CI 0.48 to 0.70, p<0.001) when compared with medical therapy.

    Only two studies15 ,26 reported isolated prosthetic valve endocarditis and analysis of these two studies showed lower odds of all-cause mortality in the early surgical group (at 20 days or less) when compared with conventional therapy without statistical significance (OR 0.80, 95% CI 0.48 to 1.34, p=0.404).

    We did a subanalysis of studies to see if there was any trend in mortality over the past 5 years (see online supplementary figure S2) and OR was in favour of early surgery in studies both before (OR 0.55, 95% CI 0.40 to 0.75, p<0.001) and after (OR 0.66, 95% CI 0.52 to 0.85, p=0.001) the year 2010.

    All-cause mortality of the propensity-adjusted groups

    We performed an analysis with studies reporting outcomes using a propensity-matched model. A total of 11 studies were included and propensity-matched variables are mentioned in online supplementary appendix 2. Again, all-cause mortality was significantly lower with early surgical intervention when compared with conventional therapy (OR 0.41, 95% CI 0.31 to 0.54, p<0.001) (figure 4). Heterogeneity was moderate (I2=41%). Sensitivity analysis performed by excluding study27 that contributed to maximum strength still showed lower mortality with early surgery (OR 0.40, 95% CI 0.29 to 0.55, p<0.001). To eliminate bias, we also separated the studies according to study type (see online supplementary figure S3). Funnel plot of the included propensity-matched studies is shown in online supplementary figure S4.

    Figure 4
    Figure 4

    Comparison of all-cause mortality between early surgery (at 20 days or less) and conventional therapy (late surgery at >20 days or medical therapy) in propensity-matched studies.

    In the subgroup of studies wherein surgery was performed at 7 days or less, odds of mortality were again in favour of the early surgical group (OR 0.30, 95% CI 0.16 to 0.54, p<0.001) (figure 5A) when compared with conventional therapy. There was a similar mortality benefit in the group that underwent surgery between 8 and 20 days (OR 0.51, 95% CI 0.35 to 0.72, p<0.001) when compared with conventional therapy (figure 5B). In the next subgroup comparing early surgery at 20 days or less with (1) medical therapy and (2) late surgery (>20 days) (figure 5C, D), odds were in favour of early surgery when compared with both late surgery (OR 0.35, 95% CI 0.16 to 0.76, p=0.008) and medical therapy (OR 0.42, 95% CI 0.31 to 0.57, p<0.001).

    Figure 5
    Figure 5

    (A) Comparison of all-cause mortality between early surgery at 7 days or less and conventional therapy in propensity-matched studies; (B) comparison of all-cause mortality between early surgery at 8–20 days and conventional therapy in propensity-matched studies; (C) comparison of all-cause mortality between early surgery (at 20 days or less) and late surgery (at >20 days) in propensity-matched studies; (D) comparison of all-cause mortality between early surgery (at 20 days or less) and medical therapy in propensity-matched studies.

    Subgroup analysis of studies done before and after the year 2010 (see online supplementary figure S5) showed odds favouring early surgical group (OR 0.36, 95% CI 0.25 to 0.52, p<0.001 and OR 0.49, 95% CI 0.35 to 0.67, p<0.001, respectively). We also did an analysis excluding studies that mentioned early surgery at ‘initial hospitalisation’ without mentioning the exact timing and that yielded similar OR favouring early surgery (OR 0.62, 95% CI 0.50 to 0.78, p<0.001) (see online supplementary figure S6).

    In-hospital mortality

    A total of 11 studies reported the incidence of in-hospital mortality and there was no significant difference between the early surgery and conventional therapy (OR 0.82, 95% CI 0.58 to 1.16, p=0.270) (figure 6A). Sensitivity analysis with one study27 removed (that had the maximum strength) yielded similar results with no significant difference in in-hospital mortality (OR 0.75, 95% CI 0.49 to 1.46, p=0.198).

    Figure 6
    Figure 6

    (A) Comparison of in-hospital mortality between early surgery (at 20 days or less) and conventional therapy; (B) comparison of in-hospital embolic events between early surgery (at 20 days or less) and conventional therapy; (C) comparison of heart failure between early surgery (at 20 days or less) and conventional therapy; (D) comparison of recurrence of endocarditis between early surgery (at 20 days or less) and conventional therapy.

    In-hospital embolic events

    Embolic events were reported in three of the included studies (figure 6B). There was no significant difference in the risk of embolic events between the early surgery versus conventional therapy (OR 0.15, 95% CI 0.01 to 1.90, p=0.142). Heterogeneity was high among the included studies (I2=77%).

    Heart failure

    The incidence of heart failure was reported in only three studies (figure 6C). Pooled OR did not show a significant difference between early intervention and conventional therapy (OR 1.78, 95% CI 0.47 to 6.74, p=0.394). Heterogeneity was moderate (I2=36%).

    Recurrence of endocarditis at follow-up

    Five studies reported recurrence of endocarditis at follow-up (figure 6D) and heterogeneity was moderate among the included studies (I2=32%). There was a higher trend of relapse with early surgery without statistical significance (OR 1.64, 95% CI 0.74 to 3.61, p=0.219).

    Discussion

    IE is associated with high mortality and morbidity. The in-hospital mortality rates vary from 15% to 20% and the 1-year mortality is estimated to be in the range of 30%–40%.3 ,31 Evidence regarding mortality benefit from early surgical intervention is derived from a limited number of studies and the approach is not without controversy. A meta-analysis32 from 2 years ago suggested the likelihood of benefit from early surgery in reducing mortality in IE but included a limited number of studies and did not comprehensively assess the timing of surgical intervention and outcomes.

    There has been no consensus on the optimal timing of early surgery for IE. Guidelines recommend early surgery during initial hospitalisation before completion of antibiotics.3

    In our study, all-cause mortality was significantly lower with early surgery compared with conventional therapy, both in unmatched and in propensity-matched groups. The results remained unaltered after exclusion-sensitivity analyses. Subgroup analysis of propensity-matched studies showed lower odds of mortality in the group that underwent surgery at 7 days or less (OR 0.30, 95% CI 0.16 to 0.54) when compared with surgery at 8–20 days (OR 0.51, 95% CI 0.35 to 0.72); however, the CIs overlap. The above observations were made in the propensity-matched groups; in the unmatched groups, the OR of mortality in the two subgroups of surgical intervention was similar and better than conventional therapy. These findings of lower risk of mortality with early surgery are in contrast to the belief that inflammatory state increases the surgical mortality. Hill et al17 in their study observed a fourfold increase in mortality in patients operated within 7 days of diagnosis of IE and was attributed to severity of IE rather than timing of surgery.

    When comparing early surgery to late surgery with medical therapy in the propensity-matched groups, the OR was significantly lower for early surgery indicating both late surgery and medical therapy are inferior to early surgery. To evaluate the effect of duration of follow-up on all-cause mortality, we divided studies into shorter (≤1 year) and longer follow-up studies (>1 year) and performed an exclusion-sensitivity analysis, which did not alter the results (see online supplementary figure S7A,B). We attempted a meta-regression of follow-up time and timing of surgery on all-cause mortality (see online supplementary figure S8A,B) and the coefficient of regression was −0.006 (p=0.21) and 0.006 (p=0.61), respectively, and so were non-significant. However, given the heterogeneous comparators used and the variable timings reported by the original investigators of the included studies, the meta-regression analysis cannot appropriately assess the impact of either follow-up time or timing of surgery on all-cause mortality.

    Embolisation33 and heart failure34 ,35 are the key contributors to mortality in IE5 and so, mortality benefit from early surgical intervention is likely due to attenuation of these events.

    In the study by Kang et al,5 embolic event rates were significantly lower with early surgery (0% vs 21%, p=0.005). All embolic events in the conservative management group occurred within 6 weeks. Longitudinal follow-up at 6 months showed mortality benefit in the surgical group. Similarly, in our study, there was a trend towards lower risk of embolic events in the early surgical group (OR 0.15, 95% CI 0.01 to 1.90) without statistical significance. Only 3 of the 21 studies reported embolic events and so significant inferences could not be drawn. Further, improvements in surgical techniques, operator experience and intensive postoperative care are likely factors in decreasing mortality in patients who undergo early surgery for IE. This could be seen from the consistently favourable odds over the past 5 years (see online supplementary figure S5).

    In our study, the in-hospital mortality was not different between early surgery and conservative therapy similar to Kang et al5 (3% vs 3%, p=1.00). We therefore think the mortality benefit is partially accrued from attenuation of embolic events. In the International collaboration of endocarditis (ICE) prospective cohort study, it was shown that the highest risk of embolism and death occured during the first week after diagnosis of IE.30 The stroke rate was 4.82/1000 patient-days within the first week; this rate dropped by 65% during the second week. We therefore believe the lower mortality in patients with surgical intervention within the first week as observed in our study could partially be secondary to the decreased embolic events following surgical intervention. Thus it could be seen that early surgery is beneficial in improving long-term survival without increasing short-term mortality.

    Severity of heart failure has been shown to increase mortality in IE.34 ,35 Hill et al17 suggested that surgical intervention within 7 days likely decreased the incidence of heart failure. This may be another reason for lower mortality in our subgroup of patients who had surgery within 7 days. Though there was an increasing trend of heart failure at follow-up with early surgical intervention, it was statistically non-significant. Also, only three studies reported the incidence of heart failure and so strong conclusions could not be derived from these data.

    Incidence of prosthetic valve endocarditis ranges between 0.1% and 2.3% per year.36 There is a lack of consensus in the management of patients with prosthetic valve endocarditis.3 Only two studies compared early surgery with conventional therapy in isolated prosthetic valve infections.15 ,26 Pooled OR showed mortality benefit in the early surgical group without statistical significance. Since it is difficult to draw conclusions based on these two studies, further randomised trials involving prosthetic valves are needed.

    There was a higher trend of recurrence of endocarditis in the early surgical group, without statistical significance. This trend was largely influenced by the study by Thuny et al,24 in which 8% of the patients, who underwent surgery within the first week, had recurrence, compared with only 2% of the patients, who had surgery between 8 and 20 days (p=0.02).24

    A recent report by Chu et al29 showed that 24% of patients with strong indications for surgery did not undergo surgical intervention and patients with Staphylococcus aureus infection had a lower likelihood of being operated. Infection with S aureus has been shown to be independently associated with neurological events.37 ,38 It could be seen that S aureus was the major pathogen in a majority of our studies (see online supplementary table S1) and so early surgical intervention should be strongly recommended for these patients. Propensity matching was done according to various micro-organisms (see online supplementary appendix S2) and all-cause mortality was better in the early surgical group. We therefore think that early surgery may be beneficial in patients with various micro-organisms including S aureus.

    The recent 2015 ESC guidelines (Class-II, level of evidence-B)4 and the 2014 AHA/ACC guidelines (Class-I, level of evidence-B)3 recommend the timing of surgical intervention be decided by a consensus of a multidisciplinary team involving cardiologists, cardiothoracic surgeons, and infectious disease specialists to reduce bias and provide best practices in patients with IE. We agree with the above recommendations and think that the Endocarditis team approach is the best strategy in decreasing mortality and improving patient outcomes. One of the recommendations is to perform early surgery for persistent bacteraemia/fever for more than 5–7 days after initiation of antibiotic therapy (Class I, level of evidence B).3 According to our analysis done from the available evidence, we suggest early surgery at 7 days or less from diagnosis carries mortality benefit on a long term. Our analysis provides the most comprehensive summary of the available literature to date until further randomised controlled trials become available.

    Limitations

    Our study has several limitations. First, our study was comprised mainly of observational studies; and is therefore subject to the individual limitations of the included studies. Secondly, all outcomes were not uniformly reported in the included studies, so heterogeneity varied across different outcomes. Thirdly, individual studies mentioning prosthetic valve endocarditis, embolic events and heart failure were limited in number. Although propensity matching (see online supplementary appendix S2) increased the strength of the study, sample sizes of the propensity matched studies were limited and there were inter-study variations in matching which could affect the results. We tried to overcome some of these limitations by performing an exclusion sensitivity analysis. In the absence of many randomised controlled trials, a meta-analysis like this provides evidence that is less prone to type-1 error. Finally, baseline characteristics, surgical procedure, timing of procedure and follow-up times varied across different studies.

    Conclusion

    Results of this meta-analysis suggest that early surgical intervention in IE increases survival when compared with conservative management and/or delayed surgery. Our results further lend support to the existing guidelines for early surgical intervention in IE and emphasise the need to optimise the timing of surgery. Further randomised studies are needed to validate these findings.

    Key messages

    What is already known on this subject?

    • Infective endocarditis is associated with high morbidity and mortality and the timing of surgery remains controversial.

    What might this study add?

    • Our comprehensive systematic review and meta-analysis suggests that early surgical intervention for infective endocarditis reduces all-cause mortality when compared with either delayed surgery or medical therapy. Further our study assesses subgroup analysis of very early surgery and suggests that this group of patients is more likely to have mortality benefit.

    How might this impact on clinical practice?

    • The observations made in our study are in contrast to the current clinical practice of delayed surgery in infective endocarditis. Based on the mortality benefits observed in our study, medical practitioners may reconsider early surgery in patients presenting with infective endocarditis.

    Acknowledgments

    The authors would like to acknowledge Mr Baskaran Krishnamurthy for reviewing the manuscript for grammatical corrections.

    References

    1. ↵
      1. Netzer ROM,
      2. Altwegg SC,
      3. Zollinger E, et al
      . Infective endocarditis: determinants of long term outcome. Heart 2002;88:61–66. doi:10.1001/archinternmed.2008.603
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Mylonakis E,
      2. Calderwood SB
      . Infective endocarditis in adults. N Engl J Med 2001;345:1318–30. doi:10.1056/NEJMra010082
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Nishimura RA,
      2. Otto CM,
      3. Bonow RO, et al
      . 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association task force on practice guidelines. Circulation 2014;129:e521–643. doi:10.1161/CIR.0000000000000031
      OpenUrlFREE Full Text
    4. ↵
      1. Habib G,
      2. Lancellotti P,
      3. Antunes MJ, et al.
      , Authors/Task Force Members. 2015 ESC guidelines for the management of infective endocarditis: the task force for the management of infective endocarditis of the European Society of Cardiology (ESC) endorsed by: European association for cardio-thoracic surgery (EACTS), the european association of nuclear medicine (EANM). Eur Heart J 2015;36:3075–128. doi:10.1093/eurheartj/ehv319
      OpenUrlFREE Full Text
    5. ↵
      1. Kang DH,
      2. Kim YJ,
      3. Kim SH, et al
      . Early surgery versus conventional treatment for infective endocarditis. N Engl J Med. 2012;366:2466–73. doi:10.1056/NEJMoa1112843
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Stroup DF,
      2. Berlin JA,
      3. Morton SC, et al
      . Meta-analysis of observational studies in epidemiology: a proposal for reporting. meta-analysis of observational studies in epidemiology (MOOSE) group. JAMA 2000;283:2008–12. doi:10.1001/jama.283.15.2008
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Moher D,
      2. Liberati A,
      3. Tetzlaff J, et al.
      , PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:b2535. doi:10.1136/bmj.b2535
      OpenUrlFREE Full Text
    8. ↵
      1. Wells G,
      2. Shea B,
      3. O'Connell D, et al
      . The newcastle-ottawa scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp (accessed 6 Jun 2014).
    9. ↵
      1. Borenstein M,
      2. Hedges LV,
      3. Higgins JPT, et al
      . eds Identifying and quantifying heterogeneity. In: Introduction to meta-analysis. John Wiley & Sons, 2009:107–25.
    10. ↵
      1. Olaison L,
      2. Hogevik H,
      3. Myken P, et al
      . Early surgery in infective endocarditis. QJM 1996;89:267–78. doi:10.1093/qjmed/89.4.267
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Bishara J,
      2. Leibovici L,
      3. Gartman-Israel D, et al
      . Long-term outcome of infective endocarditis: the impact of early surgical intervention. Clin Infect Dis 2001;33:1636–43. doi:10.1086/323785
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Vikram HR,
      2. Buenconsejo J,
      3. Hasbun R, et al
      . Impact of valve surgery on 6-month mortality in adults with complicated, left-sided native valve endocarditis: a propensity analysis. JAMA 2003;290:3207–14. doi:10.1001/jama.290.24.3207
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Mourvillier B,
      2. Trouillet JL,
      3. Timsit JF, et al
      . Infective endocarditis in the intensive care unit: clinical spectrum and prognostic factors in 228 consecutive patients. Intensive Care Med 2004;30:2046–52. doi:10.1007/s00134-004-2436-9
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Cabell CH,
      2. Abrutyn E,
      3. Fowler VG Jr., et al
      . Use of surgery in patients with native valve infective endocarditis: results from the international collaboration on endocarditis merged database. Am Heart J 2005;150:1092–8. doi:10.1016/j.ahj.2005.03.057
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Wang A,
      2. Pappas P,
      3. Anstrom KJ, et al
      . The use and effect of surgical therapy for prosthetic valve infective endocarditis: a propensity analysis of a multicenter, international cohort. Am Heart J 2005;150:1086–91. doi:10.1016/j.ahj.2005.01.023
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Aksoy O,
      2. Sexton DJ,
      3. Wang A, et al
      . Early surgery in patients with infective endocarditis: a propensity score analysis. Clin Infect Dis 2007;44:364–72. doi:10.1086/510583
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Hill EE,
      2. Herregods MC,
      3. Vanderschueren S, et al
      . Outcome of patients requiring valve surgery during active infective endocarditis. Ann Thorac Surg 2008;85:1564–9. doi:10.1016/j.athoracsur.2008.02.014
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Nadji G,
      2. Goissen T,
      3. Brahim A, et al
      . Impact of early surgery on 6-month outcome in acute infective endocarditis. Int J Cardiol 2008;129:227–32. doi:10.1016/j.ijcard.2007.07.087
      OpenUrlCrossRefPubMed
    19. ↵
      1. Sy RW,
      2. Bannon PG,
      3. Bayfield MS, et al
      . Survivor treatment selection bias and outcomes research: a case study of surgery in infective endocarditis. Circ Cardiovasc Qual Outcomes 2009;2:469–74. doi:10.1161/CIRCOUTCOMES.109.857938
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Cooper HA,
      2. Thompson EC,
      3. Laureno R, et al
      . Subclinical brain embolization in left-sided infective endocarditis: results from the evaluation by MRI of the brains of patients with left-sided intracardiac solid masses (EMBOLISM) pilot study. Circulation 2009;120:585–91. doi:10.1161/CIRCULATIONAHA.108.834432
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Kim DH,
      2. Kang DH,
      3. Lee MZ, et al
      . Impact of early surgery on embolic events in patients with infective endocarditis. Circulation 2010;122(11 Suppl):S17–22. doi:10.1161/CIRCULATIONAHA.109.927665
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Lalani T,
      2. Cabell CH,
      3. Benjamin DK, et al
      . Analysis of the impact of early surgery on in-hospital mortality of native valve endocarditis: use of propensity score and instrumental variable methods to adjust for treatment-selection bias. Circulation 2010;121:1005–13. doi:10.1161/CIRCULATIONAHA.109.864488
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Bannay A,
      2. Hoen B,
      3. Duval X, et al
      . The impact of valve surgery on short- and long-term mortality in left-sided infective endocarditis: do differences in methodological approaches explain previous conflicting results? Eur Heart J 2011;32:2003–15. doi:10.1093/eurheartj/ehp008
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Thuny F,
      2. Beurtheret S,
      3. Mancini J, et al
      . The timing of surgery influences mortality and morbidity in adults with severe complicated infective endocarditis: a propensity analysis. Eur Heart J 2011;32:2027–33. doi:10.1093/eurheartj/ehp089
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Funakoshi S,
      2. Kaji S,
      3. Yamamuro A, et al
      . Impact of early surgery in the active phase on long-term outcomes in left-sided native valve infective endocarditis. J Thorac Cardiovasc Surg 2011;142:836–42.e1. doi:10.1016/j.jtcvs.2011.01.040
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Lalani T,
      2. Chu VH,
      3. Park LP, et al
      . In-hospital and 1-year mortality in patients undergoing early surgery for prosthetic valve endocarditis. JAMA Intern Med 2013;173:1495–504. doi:10.1001/jamainternmed.2013.8203
      OpenUrlCrossRefPubMed
    27. ↵
      1. Gálvez-Acebal J,
      2. Almendro-Delia M,
      3. Ruiz J, et al
      . Influence of early surgical treatment on the prognosis of left-sided infective endocarditis: a multicenter cohort study. Mayo Clin Proc 2014;89:1397–405. doi:10.1016/j.mayocp.2014.06.021
      OpenUrlCrossRefPubMed
    28. ↵
      1. Martinez-Sellés M,
      2. Muñoz P,
      3. Arnáiz A, et al
      . Valve surgery in active infective endocarditis: a simple score to predict in-hospital prognosis. Int J Cardiol 2014;175:133–7. doi:10.1016/j.ijcard.2014.04.266
      OpenUrlCrossRefPubMed
    29. ↵
      1. Chu VH,
      2. Park LP,
      3. Athan E, et al
      . Association between surgical indications, operative risk, and clinical outcome in infective endocarditis: a prospective study from the international collaboration on endocarditis. Circulation 2015;131:131–40. doi:10.1161/CIRCULATIONAHA.114.012461
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Dickerman SA,
      2. Abrutyn E,
      3. Barsic B, et al
      . The relationship between the initiation of antimicrobial therapy and the incidence of stroke in infective endocarditis: an analysis from the ICE prospective cohort study (ICE-PCS). Am Heart J 2007;154:1086–94. doi:10.1016/j.ahj.2007.07.023
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Prendergast BD,
      2. Tornos P
      . Surgery for infective endocarditis: who and when? Circulation 2010;121:1141–52. doi:10.1161/CIRCULATIONAHA.108.773598
      OpenUrlFREE Full Text
    32. ↵
      1. Chatterjee S,
      2. Sardar P
      . Early surgery reduces mortality in patients with infective endocarditis: insight from a meta-analysis. Int J Cardiol 2013;168:3094–7. doi:10.1016/j.ijcard.2013.04.078
      OpenUrlCrossRefPubMed
    33. ↵
      1. García-Cabrera E,
      2. Fernández-Hidalgo N,
      3. Almirante B, et al
      . Neurological complications of infective endocarditis: Risk factors, outcome, and impact of cardiac surgery: a multicenter observational study. Circulation 2013;127:2272–84. doi:10.1161/CIRCULATIONAHA.112.000813
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. López J,
      2. Sevilla T,
      3. Vilacosta I, et al
      . Clinical significance of congestive heart failure in prosthetic valve endocarditis. A multicenter study with 257 patients. Rev Esp Cardiol (Engl Ed) 2013;66:384–90. doi:10.1016/j.rec.2012.10.022
      OpenUrlCrossRefPubMed
    35. ↵
      1. Kiefer T,
      2. Park L,
      3. Tribouilloy C, et al
      . Association between valvular surgery and mortality among patients with infective endocarditis complicated by heart failure. JAMA 2011;306:2239–47. doi:10.1001/jama.2011.1701
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Blackstone EH,
      2. Kirklin JW
      . Death and other time-related events after valve replacement. Circulation 1985;72:753–67. doi:10.1161/01.CIR.72.4.753
      OpenUrlAbstract/FREE Full Text
    37. ↵
      1. Heiro M,
      2. Nikoskelainen J,
      3. Engblom E, et al
      . Neurologic manifestations of infective endocarditis: a 17-year experience in a teaching hospital in Finland. Arch Intern Med 2000;160:2781–7. doi:10.1001/archinte.160.18.2781
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Vilacosta I,
      2. Graupner C,
      3. San Roman J.A, et al
      . Risk of embolization after institution of antibiotic therapy for infective endocarditis. J Am Coll Cardiol 2002;39:1489–95. doi:10.1093/eurheartj/ehm005
      OpenUrlCrossRefPubMedWeb of Science
    View Abstract

    Supplementary materials

    Footnotes

    • Contributors Data access responsibility and analysis: all authors had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study and concept design: MAN, RV, and RMS. Acquisition of data: MAN, TMH, and GM. Analysis or interpretation of data, drafting of the manuscript, administrative, technical or material support: all authors.

    • Competing interests CJD: Received grants from Cubist, Durata Therapeutics and Forest labs. RMS: Research funding— Edwards Lifesciences, St. Jude Medical and Sorin Group. Current technology licensing agreements— Sorin Group. Principle Investigator—National PI FDA IDE PERCEVAL Trial—Sorin Group. Co-Investigator— PARTNER II, COAPT, SURTAVI and PORTICO. No personal financial relationships with industry.

    • Provenance and peer review Not commissioned; externally peer reviewed.