Xuxi Zhang

Table 4 Concurrent validity of the TFI and its three domains (n=2250) (Continued) Adverse outcomes (measures) Screening Population Cut ‐ off Point * Sensitivity Specificity AUC (95% CI) † Feeling lonely (short ‐ JG) Full TFI Total ≥ 6 0.66 0.76 0.79(0.77, 0.81) Spain ≥ 6 0.59 0.78 0.75(0.71, 0.80) Greece ≥ 7 0.55 0.81 0.74(0.69, 0.79) Croatia ≥ 8 0.54 0.81 0.73(0.68, 0.77) The Netherlands ≥ 5 0.73 0.77 0.84(0.80, 0.88) The United Kingdom ≥ 5 0.79 0.69 0.79(0.75, 0.84) Social domain Total ≥ 2 0.60 0.79 0.74(0.72, 0.76) Spain ≥ 2 0.61 0.78 0.74(0.70, 0.79) Greece ≥ 2 0.73 0.64 0.71(0.66, 0.77) Croatia ≥ 2 0.65 0.69 0.69(0.64, 0.75) The Netherlands ≥ 1 0.75 0.64 0.73(0.68, 0.79) The United Kingdom ≥ 1 0.86 0.55 0.76(0.71, 0.80) AUC, area under ROC curve; ROC, receiver operating characteristic; CI, confidential interval; TFI, Tilburg Frailty Indicator; SHARE ‐ FI, Survey of Health, Ageing and Retirement in Europe ‐ Frailty Instrument; GARS, 18 ‐ item Groningen Activity Restriction Scale; GALI, Global Activity Limitation Index; MHI ‐ 5, full 5 ‐ item mental well ‐ being scale the 36 ‐ Item Short Form Survey; short ‐ JG, 6 ‐ item version of the De Jong Gierveld loneliness scale Missing items: SHARE ‐ FI = 36; GARS = 4; GALI = 8; MHI ‐ 5 = 22; short ‐ JG = 27 * The Youden index was adopted as the criterion for selecting the optimum cut ‐ off point; if more than one cut ‐ off points had the maximum value, all potential cut ‐ off points as well as corresponding sensitivity and specificity were provided. † 0.7 ≤ AUC<0.8 is considered acceptable concurrent validity; 0.8 ≤ AUC<0.9 excellent; AUC ≥ 0.9 outstanding; The value of AUC ≥ 0.7 in bold. 4 91 Reliability and validity of the TFI