Beginning in June 2018, serologic screening for Q fever was ordered for all patients admitted for TAVI at Rambam Health Care Campus, a 960-bed primary and tertiary university-affiliated hospital in northern Israel. We tested serum samples for C. burnetii phase 2 IgM and phase I or phase II IgG by using ELISA (Institute Virion/Serion GmbH, https://www.virion-serion.de). For samples that tested positive, we then conducted an indirect immunofluorescence assay (IFA) for confirmation and titer determination. We performed the IFA locally using a commercial kit (Focus Diagnostics, https://www.focusdx.com) or an in-house test at The National Reference Laboratory for Rickettsiosis (Nes Ziona, Israel). An infectious diseases specialist evaluated patients with positive IgG for C. burnettii chronic infection. IE was diagnosed according to the modified Duke criteria (6) or the Dutch consensus guidelines of chronic Q fever infection (7) with an IFA phase I IgG of >800. Patients began treatment and follow-up was conducted at the infectious disease and cardiology outpatient clinics. Diagnostic testing was performed as a part of a clinical routine, and anonymous data collection was approved by the hospital's ethics committee with a waiver of informed consent.

During June 1, 2018–May 31, 2020, a total of 197 TAVI procedures were performed at Rambam Health Care Campus. Serologic testing for Q fever was conducted in 155 patients. Nine patients tested positive for >1 Q fever IgG by ELISA: 7 had phase I IgG and 2 patients had only phase II IgG. On IFA, 4 patients (2.6%) had a phase I IgG titer of >800 and were further evaluated for Q fever IE (Table). All 4 patients had underlying conditions, but none had fever or vegetations on echocardiography. None of the patients had a specific high-risk exposure for Q fever. We recommended treatment with doxycycline and hydroxychloroquine for >24 months (as recommended for Q fever IE in the presence of prosthetic valve). In 3 of 4 patients, treatment was modified to an alternative regimen because of intolerance or side effects. We did not perform fluorodeoxyglucose positron emission tomography-computed tomography for diagnosis, because it would not have led to a change in management. Patient 2 underwent fluorodeoxyglucose positron emission tomography–computed tomography 2 months before TAVI as part of lymphoma follow-up; it showed no evidence of pathologic uptake in the valve or elsewhere. As of the last follow-up visit (median 12 months, range 8–18 months), all 4 patients had preserved prosthetic valve function, and none experienced symptomatic Q fever infection. One patient reported severe fatigue, likely related to underlying scleroderma.

During a 2-year period of routine serologic screening for Q fever among patients undergoing TAVI, we identified 4 case-patients with Q fever IE, affecting 2.6% of patients screened. None of the 4 case-patients experienced fever or echocardiographic findings that were suggestive of IE.

Diagnosing Q fever IE can be challenging, especially in the absence of tissue samples, as in the case of patients undergoing TAVI. Several studies have highlighted the difficulties of the diagnosis of Q fever IE (Appendix Table 1). The diagnostic criteria used in the absence of tissue samples are based on the modified Duke criteria (6), the Dutch consensus guidelines for chronic Q fever (7), and the recently revised definition of “persistent C. burnettii infection” by Melenotte et al. (8) (Appendix Table 2). For definitive diagnosis, all 3 definitions are based mainly on serologic tests and echocardiography, PET, or CT findings to prove valve infection. Both imaging modalities have poor sensitivity in the case of C. burnettii IE (9–11). The alternative minor diagnostic criteria consist also of infrequent findings, such as embolic and immunologic phenomena. We recommended treatment for patients with possible or probable IE (Table), recognizing the significant consequences of a delayed diagnosis and treatment of prosthetic Q fever IE among patients at very high risk for surgery a priori.

In a study conducted in 2 centers in the United Kingdom, routine serologic screening for Q fever before valve surgery was performed in 139 patients. In this low-endemicity setting, no patient with Q fever IE was identified (12). In our study conducted in a Q fever–endemic region, the yield of such a strategy seems clinically significant. The incidence of Q fever in Israel according to reported cases to the Ministry of Health is ≈2.2/100,000 population (https://www.health.gov.il/UnitsOffice/HD/PH/epidemiology/Pages/epidemiology_report.aspx). In comparison, data from countries in the European Union from 2018 showed the highest incidence was 0.7/100,000 population in Spain. An alternative indicator of Q fever endemicity is the percentage of IE caused by Q fever out of all IE cases. According to the International Collaboration on Endocarditis registry data, C. burnettii was responsible for almost 1% of all IE cases in 25 countries (13). This rate reaches almost 5% in Q fever–endemic regions, such as southern France (14). A similar rate was observed at our hospital; Q fever IE was diagnosed in 5 (5.3%) of 95 cases of definitive IE during 2013–2016, according to local data from a prospective registry.

The primary limitation of our study is that, as a single-center study, it reflects the epidemiology of a limited geographic area. The short-term follow-up of patients with Q fever IE does not enable a description of the long-term benefit of our strategy. We did not evaluate the cost-effectiveness of our surveillance strategy. In addition, we might have missed cases of Q fever IE by conducting serologic screening only, since Q fever IE with low phase I IgG titers (<800) (9) or even negative serologic results (15) has been described. Nevertheless, as a screening strategy, serologic testing seems to be sufficient. Early diagnosis and appropriate treatment as soon as possible after prosthetic valve implantation contributed substantially to preserving valve function and prevented potential ongoing infection. Therefore, we suggest screening for Q fever in TAVI patients in settings in which Q fever incidence is >0.5 per 100,000 (nationally or in Q fever–endemic regions within countries), after Q fever outbreaks regardless of baseline incidence, or in places in which Q fever causes >2% of all cases of IE.