QT prolongation is associated with increased mortality in different types of patients [1, 2]. QT prolongation is common in critically ill patients [3], and the association between heart rate corrected QT (QTc) and outcome in critically ill patients has raised broad interests most recently [4]. However, the prevalence of new-onset QT prolongation and its significance in these patients was not well studied yet.

Here, we prospectively recruited 505 consecutive ICU patients without known previous QT prolongation to evaluate the risk factors for new-onset QT prolongation and the prognostic value of QTc calculated by different methods. The baseline clinical and laboratory characteristics of subjects were shown in Table 1. The mean Bazett QT interval was 413.6 ± 33.8 ms. New-onset QT prolongation occurred in 99 patients (19.6%). This occurrence is about 200-fold higher than that in the general population [5]. Intriguingly, the occurrence of nonthyroidal illness syndrome (NTIS) is significantly higher in patients with QT prolongation than those without (Table 1), indicating that NTIS might be a risk factor of QT prolongation. Indeed, multivariate linear regression showed that QTc was independently associated with NTIS, heart rate, level of serum potassium, gender, and estimated glomerular filtration rate (eGFR).

Table 1 Clinical and laboratory characteristics of subjects
Full size table

There was a significantly graded increase in mortality rate across increasing QTc quintile (p = 0.004) (Fig. 1a). The overall mortality rate in patients with a new-onset QTc prolongation is more than two times higher than those patients without (22.2% vs 9.6, OR = 2.69, p = 0.001) (Fig. 1b). Multivariate logistic regression showed that QT prolongation is still independently associated with ICU mortality even after adjusted for age and gender (p = 0.001, 95% C.I., 1.51–4.79). However, QT prolongation is no longer a predictor of ICU mortality if APACHE-II score was further adjusted (p = 0.329), likely due to that QTc itself is strongly associated with APACHE-II score (r = − 0.235, p < 0.001).

Fig. 1
figure 1

a ICU mortality rates among patients with different QTc quintile. *Linear-by-Linear Association by chi-square test. b Overall ICU mortality rates among patients with or without QT prolongation

Full size image

As Bazett’s formula can over-correct QT at high heart rates and under-correct it at low heart rates, we then evaluated the prognostic value of QTc calculated using additional formulas including Fridericia’s, Framingham’s, and Hodges’s. We found that patients in quintile 5 have significantly higher mortality than patients in the combination of quintiles 1–4 regardless of which formula was used (all p < 0.05).

In summary, QT prolongation determined by baseline ECG can serve as a novel indicator of the severity of illness in critically ill patients. NTIS is a new risk factor of QT prolongation in critically ill patients.