Association of Time to Rapid Response Team Activation With Patient Outcomes Using a Range of Physiologic Deterioration Thresholds: Erratum

In the following article in the November 2022 compendium, the following errors were made.

In the Conclusions section of the abstract, the following sentence should have been:

This relationship appeared most marked when using a CART score greater than or equal to 20 threshold from which to measure time to RRT activation.

In the Key Points box on page 2, the final sentence in the Findings section should have been:

This relationship appeared most marked when using a Cardiac Arrest Risk Triage score greater than or equal to 20 threshold from which to measure time to RRT activation.

On page 4, in the second paragraph of the Main Results section, the sentence should have read:

This was most apparent when using CART thresholds of greater than or equal to 16 and greater than or equal to 20 (Fig. 3, B and C) and least apparent when using a CART threshold of greater than or equal to 24.

In the legend for Figure 3, the first sentence should have read:

Predicted probabilities of 7-d death or discharge to hospice with increasing time to rapid response team (RRT) activation in RRT activation after Cardiac Arrest Risk Triage (CART) score of greater than or equal to 12 (green, A), greater than or equal to 16 (yellow, B), greater than or equal to 20 (orange, C), or greater than or equal to 24 (red, D).

For the final two paragraphs of the Main Results section, they should have read:

We visually inspected the primary analysis fractional polynomial models to identify inflection points that could be used to categorize time to RRT activation. We prioritized evident inflection points from the model using a CART threshold greater than or equal to 20, since it showed the most marked relationship between increasing time to RRT activation and mortality, although the model using a CART threshold of greater than or equal to 16 demonstrated inflection points at similar time points. Table 2 shows the resultant categories and their adjusted odds of 7-day mortality, presented for each CART threshold. We repeated this using CART greater than or equal to 16 and CART greater than or equal to 12 fractional polynomial models to inform time to RRT activation categories (Supplemental Table 1, http://links.lww.com/CCX/B83). We found an association of increased odds of 7-day mortality with increasing categories of time to RRT activation after meeting deterioration criteria. This association was not observed when using a CART threshold of greater than or equal to 24 to define deterioration and was consistent when time to RRT activation of less than 1 hour or less than 2 hours was used as the reference time category. A similar association was observed between increasing categories of time to RRT activation and 30-day mortality (Supplemental Table 2, http://links.lww.com/CCX/B83).

Of the 1,607 patients with a CART score greater than or equal to 12 in the 24 hours prior to RRT activation, 457 (28%) met the threshold of CART greater than or equal to 20 at the same time, 432 (27%) progressed to a CART score greater than or equal to 20 after they reached CART greater than or equal to 12, and 718 (45%) never progressed to a CART score greater than or equal to 20. Among the 432 patients who progressed to a CART score greater than or equal to 20, it took a median of 16 hours (interquartile range, 6–24 hr) for CART scores to increase from greater than or equal to 12 to greater than or equal to 20. The relationship between increasing time to RRT activation and increased mortality was most marked among patients who progressed from a CART score greater than or equal to 12 to subsequent CART score greater than or equal to 20 and was much less apparent among those who never progressed to CART greater than or equal to 20 or those who exceeded the threshold of CART greater than or equal to 12 and CART greater than 20 at the same time (Supplemental Fig. 8, http://links.lww.com/CCX/B83).

In the first paragraph of the Conclusions section, the following sentence should have read:

Additionally, there appeared to be a more marked increase in probability of mortality with shorter time to RRT activation when using higher CART thresholds to define T₀, a relationship that appeared most evident with a CART score threshold of greater than or equal to 20.

In the last sentence of the third paragraph of the Conclusions section, the following sentence should have read:

It is also possible that we were not adequately powered to detect the association of increased time to RRT activation with mortality among patients with CART scores greater than or equal to 24, as the number of patients with such deterioration represented less than one quarter of the study sample.

In the first sentence of the sixth paragraph of the Conclusions section, the following sentence should have read:

Examining trajectory in those who reached CART greater than or equal to 12, we found that patients who progressed over time to severe deterioration before RRT activation (CART ≥ 20) had a stronger time-mortality relationship than in those whose vital sign derangements did not progress.

In the second sentence of the final paragraph of the Conclusions section, the following sentence should have read:

This relationship appeared most marked when using a CART score greater than or equal to 20 threshold from which to measure time to RRT activation.