Bisphosphonates and risk of atrial fibrillation: a meta-analysis

Table 2 Summary of included case-control studies and characteristics

Study, Year, Country	Subjects	Mean age (year)	Mean duration (year)	Type of bisphosphonate	AF definition	Total AF events (n)	Variables matched and/or controlled	Quality assessment score^§
Heckbert and colleagues [7], 2008, USA	1,685 women in a commercial health care system database	73	20	Alendronate	ICD-9 codes verified by medical records +/- EKG	719	Age, HTN, calendar year, osteoporosis, and any cardiovascular disease	4/2/3
Sorensen and colleagues [19], 2008, Denmark	81,640 women in the National Registry of Patients	>75	Not reported	Alendronate Etidronate	ICD-8 and 10 codes	13586	Age, county, cardiovascular disease, renal failure, DM, HRT, fracture, cancer, alcoholism, pulmonary disease, anti-thyroid drugs, osteoporosis, steroid use, etc	3/2/3
Grosso and colleagues* [16], 2009, UK	2,195 women with AF in the UK General Practice Research Database	82	13	Alendronate Risedronate	Diagnosis codes and Read codes	2195	Age⁺	3/2/3
Levesque and colleagues [18], 2009, Canada	63,843 adults (BIS users and raloxifene users) in the Quebec health database	75	4	Alendronate Etidronate^a Risedronate^a	Not reported	2149	Age, cohort entry year, follow-up, CAD, HTN, and use of HRT	N/A

AF, atrial fibrillation; BIS, bisphosphonates; CAD, coronary artery disease; DM, diabetes mellitus; EKG, electrocardiogram; HRT, hormone replacement therapy; HTN, hypertension; ICD, International Classification of Diseases.
* self-controlled case-series. ⁺ no other adjustment not needed in self-controlled study, N/A, not applicable (only the abstract was available at the time of this study), ^a risks specific to etidronate or risedronate not used in meta-analysis, ^§ Methodologic quality assessment score based on Newcastle-Ottawa Scale [23]; selection (0 to 4), comparability (0 to 2), and outcome score (0 to 3) points with a higher score representing a better quality.

ISSN: 1478-6362