Table 2.
Characteristics of the included studies with respect to the sample used to assess diagnostic accuracy of the algorithms (i.e., internal or external validation)
| Author, year | Country | Period | Study design | Population | N of pts | Age (mean/median) | Females (%) | HF prevalence (%) | Type of reference standardh (criteria) |
|---|---|---|---|---|---|---|---|---|---|
| Dunlay, 2022 [33] | USA | 2007–2017 | Longitudinal | General | 8657 | 74.1 | 49.8 | 9.8 | RCR (ESC) |
| Vijh, 2021 [34] | Canada | 2018–2018 | Longitudinal | General, COPD | 311 | 73.2 | 63.7 | 23.2 | RCR (CCS, ESC, ACCF/AHA) |
| Xu, 2020 [35] | Canada | 2015–2015 | Cross-sectionala | Hospitalized | 2105 | 64 | 50.2 | 14.1 | RCR (ad hoc) |
| Cozzolino, 2019 [36] | Italy | 2012–2014 | Case control | Hospitalized | 203 | 81.5c,e | 53c | n.c | RCR (ESC) |
| Kaspar, 2018 [51] | Germany | 2000–2015 | Cross-sectional | Hospitalized | 1042 | 77.5d,e | 41d | 21.3 | RCR (cardiology expertise) |
| Tison, 2018 [37] | USA | 2010–2012 | Longitudinal | General | 76,254 | 52.0e | 52.9e | 2.9 | RCR (Framingham) |
| Franchini, 2018 [38] | Italy | 2011–2014 | Case control | General | 389 | 76.3% > 65 yrs | 48.1 | n.c | RCR (cardiology expertise) |
| Bosco-Lévy, 2019 [50] | France | 2014–2014 | Cross-sectional | Hospitalized | 229 | 69.1, 76.8f | 28.8 | n.c | RCR (ESC) |
| Blecker, 2016 [52] | USA | 2013–2015 | Cross-sectionalb | Hospitalized | 6549 | 60.9 | 50.8 | NA | RCR (ARIC study) |
| Schultz, 2013 [39] | Canada | 2004–2007 | Longitudinal | General | 2338 | 57.9 | 56 | 4.2 | RCR (ad hoc) |
| Alqaisi, 2009 [40] | USA | 2004–2005 | Longitudinal | General | 400 | 68e | 54 | 65.0 | RCR (Framingham) |
| Teng, 2008 [41] | Australia | 1996–2006 | Cross-sectional | Hospitalizedg | 1006 | 79.5e | 49.7 | n.c | RCR (Boston score) |
| Kümler, 2008 [53] | Denmark | 1998–1999 | Cross-sectional | Hospitalized | 3201 | 70.8e | 33.3e | 13.4 | Medical evaluation (ESC) |
| So, 2006 [42] | Canada | 1994–2004 | Cross-sectional | Hospitalized, AMI | 193 | 68.0d,e | 34.4d | 28.5 | RCR(NR) |
| Ingelsson, 2005 [43] | Sweden | 1976–2001 | Cross-sectional | Hospitalized | 2322 | NA | 0 | n.c | RCR (ESC) |
| Lee, 2005 [44] | Canada | 1997–1999 | Cross-sectional | Hospitalized | 1641 | 75.5c,e | 50.9c,e | n.c | RCR (Framingham) |
| Birman-Deych, 2005 [45] | USA | 1998–1999 | Cross-sectional | Hospitalized, AF | 23,657 | 78.8 | 55 | 46.6 | Pathology registry (NR) |
| Wilchesky, 2004 [46] | Canada | 1995–1996 | Longitudinal | General | 14,980 | NA | NA | 7.1 | RCR (primary care physician diagnosis) |
| Borzecki, 2004 [47] | USA | 1998–1999 | Longitudinal | General, veterans, HT | 1176 | NA | NA | 7.0 | RCR (attending physician diagnosis) |
| Quan, 2002 [54] | Canada | 1996–1997 | Cross-sectionala | Hospitalized | 1200 | NA | NA | 10.7 | RCR (Charlson) |
| Austin, 2002 [55] | Canada | 1996–2000 | Cross-sectional | Hospitalized | 428 | 66.5 | 39 | 9.3 | Pathology registry (attending physician diagnosis) |
| Szeto, 2002 [19] | USA | 1996–1998 | Longitudinal | General, veterans | 148 | 64 | 4 | 10.1 | RCR (NR) |
| Udris, 2001 [48] | USA | 1996–2000 | Cross-sectional | General, veterans | 2246 | 68.8 | 3 | 34.6 | RCR (ad hoc) |
| Jollis, 1993 [49] | USA | 1985–1990 | Cross-sectional | Hospitalized, CA | 12,854 | 58.8 | 33.7 | 13.9 | Research database (NR) |
Abbreviations: ACCF/AHA American College of Cardiology Foundation/American Heart Association, AF atrial fibrillation, AMI acute myocardial infarction, CA undergoing coronary angiography, CCS Canadian Cardiovascular Society, COPD chronic obstructive pulmonary disease, ESC European Society of Cardiology, HT hypertension, NA not available, n.c. not meaningful to compute, RCR review of clinical records
aQuan H. et al. Med Care 2005; 43:1130. bBlecker S. et al. J Am Coll Cardiol. 2013; 61:1259. cReported only for subjects positive to the algorithm. dData relating only to subjects used for validation were not available; data for the whole population were extracted on the rationale that subjects used for validation were a random sample of the whole population. eWeighted mean of average in reported classes. fThe first value is referred to the cohort used to evaluate Sn and Sp, the second one to a different cohort used to assess the PPV. gNonelective admission only. hReferences for the criteria and ad hoc definitions of the reference standard are detailed in Additional file 4, Supplementary Table S1