Population-Based Surveillance of Infectious Endocarditis in an Italian Region FREE

In a recent article, Murdoch and coworkers¹ report on a prospective cohort of 2781 adults with definite infectious endocarditis (IE), admitted to 58 hospitals in 25 countries from June 1, 2000, through September 1, 2005. As they acknowledge, one major limitation of the study is the “catchment area” of enrolled patients, namely centers with a particular interest in IE, typically referral centers with cardiac surgical programs. Admittedly, the study population was unlikely to be a true population-based sample, thereby limiting epidemiologic inferences. This is a common issue in most literature regarding IE.²

Based on this concern, in 2004, we started a continuous surveillance of all cases of IE in our region, Friuli-Venezia Giulia (FVG) (population 1 200 000) in Italy.³ We report herein the methods used and preliminary results.

All 29 clinical centers potentially admitting cases of IE in FVG participated in the study (Table 1). The centers comprised general medicine wards (n = 16), cardiology wards and their echocardiography service (n = 9), and wards for infectious diseases (n = 2) and cardiovascular surgery (n = 2). Prior to study start, each center obtained ethics committee clearance. To ensure commitment to study participation and notification of all cases, a yearly meeting was organized to report on and discuss study progress and results.

For all patients with suspected or confirmed IE seen at one of the participating centers (and in particular, to ensure the maximal sensitivity, at one of the echocardiography services), a “warning alarm form” was faxed to the coordinating center (Clinic of Infectious Diseases, University of Udine, Udine, Italy) and included information on patient demographics, admittance hospital and ward, and attending physician and attending cardiologist with contact information for follow-up. In the following days, a standardized, detailed case report form was completed by the attending physician, including demographic, clinical (predisposing conditions and risk factors), radiological, microbiological, laboratory, treatment (including surgery), and outcome (at the end of treatment) data. Because this was an observational study, no indication was given as to antibiotic treatment or surgical indications. Dedicated medical personnel ensured that all cases notified to the system via the warning alarm forms had a case record form completed. Case record forms were sent to the coordinating center, where they were reviewed by a panel that comprised an infectious disease specialist (P.V.) and an expert on echocardiography (L.B.) to ensure that the diagnosis of IE was consistent with modified Duke criteria.⁴ Data were subsequently entered into an electronic database created with Access (Microsoft Corp, Redmond, Washington) by the aforementioned medical personnel, who also checked data completeness and consistency and, if needed, issued queries to the relevant clinical center. Cases not residing in FVG were excluded from the present report. Cross-checking with the regional database of discharge medical records was periodically performed to ensure that all cases were notified to the system.

As of December 2008, 254 cases (241 [94.9%] with definite IE) were notified. Incidence was 4.21 cases/100 000/y, and 12.88 cases/100 000/y in the population older than 65 years. The mean (SD) age of patients was 67 (14) years, 33.5% were female, and 32.3% had prosthetic valve IE. The median time from clinical onset and diagnosis was 20 days (interquartile range, 7-35 days). Demographics, risk factors, and clinical characteristics are reported in Table 2. Interestingly, in our population the most frequent isolate was Enterococcus species (n = 48 [18.9%]), particularly in the elderly population, followed by Staphylococcus aureus (n = 46 [18.1%]) and viridans group streptococci (n = 42 [16.5%]) (Figure). However, a microbiological diagnosis was not made for 49 cases (19.3%). The median antibiotic course was 41 days (interquartile range, 28-56 days). Surgery was performed in 102 cases (40.2%), of which 74 (72.6%) were performed during antibiotic therapy. There were 52 deaths (20.5%).

We demonstrated that area-wide surveillance for IE is feasible, although requiring good intercenter cooperation and communication. We think that the epidemiological, clinical, and outcome features of this complete patient population represent the unbiased picture of IE in our area; the relatively high proportion of patients lacking microbiological diagnosis most likely depicts a common situation in the clinical practice of general and small hospitals, where the due attention to preanalytic phase of blood cultures is not paid and where antibiotic treatment is empirically initiated at hospital admission, particularly in elderly and frail patients. Possible exclusions included postmortem (autopsy) diagnoses, pediatric cases (pediatric wards were not included among the participant centers), patients with very mild cases that spontaneously resolved and did not present for medical care, patients who received a diagnosis and were treated by their general physician (in Italy, this would be a negligible proportion of patients with IE), and patients with intravenous drug abuse not reporting for medical attention, which in FVG are estimated to be few. Overestimation due to referrals from other areas is unlikely, since we excluded cases not residing in FVG.

The high incidence of Enterococcus species was striking evidence. It might be due to a high prevalence of elderly persons in our area (persons >65 years represent 22.8% of the population⁵) and to the high referral rate to medical attention among northeastern Italian populations; in fact, among the elderly population, incidence of enterococcal disease was particularly relevant. An alternative explanation would be that in our study the higher than usual proportion of cases lacking microbiological diagnoses would include cases due to organisms (like streptococci) with wide antibiotic sensitivity spectrum and therefore with antimicrobial clearance quicker than enterococci when exposed to empirical treatment.

A third possible explanation would be that other studies lacked the ability to catch cases caused by enterococci because they were less likely to be referred to the participating centers owing to their frequently mild clinical course or, on the contrary, to the presence of conditions limiting the possibility of undergoing a surgical approach (such as advanced age or comorbidities).

In fact, in our study we documented an extremely high rate of embolization (19.4% at diagnosis plus another 15.8% in the course of treatment [total, 35.2%]), which might indicate that many cases were included that other studies might not have caught. While we are preparing a full report of our experience, it would be interesting if other groups would reproduce our surveillance system in other settings, to add another tile to the global mosaic of IE epidemiology.

Correspondence: Dr Viale, Clinic of Infectious Diseases, University of Udine and Azienda Ospedaliero–Universitaria di Udine, Piazzale Santa Maria della Misericordia 15, 33100 Udine, Italy (pierluigi.viale@uniud.it).

Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Scudeller, Badano, and Viale. Acquisition of data: Scudeller, Crapis, and Pagotto. Analysis and interpretation of data: Scudeller and Viale. Drafting of the manuscript: Scudeller, Crapis, and Viale. Critical revision of the manuscript for important intellectual content: Badano, Pagotto, and Viale. Statistical analysis: Scudeller. Obtained funding: Viale. Administrative, technical, and material support: Crapis, Pagotto, and Viale. Study supervision: Badano and Viale.

Financial Disclosure: None reported.

Previous Presentation: This study was presented in part as a poster at the 10th International Symposium on Modern Concepts in Endocarditis and Cardiovascular Infections; April 26-28, 2009; Naples, Italy.

Additional Contributions: We thank all the members of “Registro Regionale Endocarditi Infettive del Friuli-Venezia Giulia” for their active contribution to study design, conduct, and comments.