by Jeyaraju, Maniraj; McCurdy, Michael T.; Levine, Andrea
R.; Devarajan, Prasad; Mazzeffi, Michael A.; Mullins, Kristin E.; Reif,
Michaella; Yim, David N.; Parrino, Christopher; Lankford, Allison S.; Chow,
Jonathan H.
Critical Care Medicine: January 2022
- Volume 50 - Issue 1 - p 50-60
Objectives:
Whole blood lactate concentration is widely used in shock
states to assess perfusion. We aimed to determine if the change in plasma renin
concentration over time would be superior to the change in lactate
concentration for predicting in-hospital mortality in hypotensive patients on
vasopressors.
Design:
Prospective, observational cohort study.
Setting:
Tertiary academic ICU.
Patients:
Adult patients on vasopressors for greater than 6 hours to
maintain a mean arterial pressure greater than or equal to 65 mm Hg during
January 2020.
Interventions:
Plasma renin concentrations were measured at enrollment and
at 24, 48, and 72 hours. Whole blood lactate measurements were performed
according to normal standard of care. Logistic regression was performed to
evaluate whether the change in renin or lactate concentration could predict
in-hospital mortality. Generalized estimating equations were used to analyze
the association between renin and lactate concentration and in-hospital
mortality. The area under the receiver operating characteristics curve was
performed to measure the discriminative ability of initial and peak renin and
lactate concentration to predict mortality. The association between renin and
lactate concentration above the upper limit of normal at each timepoint with
in-hospital mortality was also examined.
Measurements and Main Results:
The study included 197 renin and 148 lactate samples
obtained from 53 patients. The slope of the natural log (ln) of renin
concentration was independently associated with mortality (adjusted odds ratio,
10.35; 95% CI, 1.40–76.34; p = 0.022), but the slope of ln-lactate
concentration was not (adjusted odds ratio, 4.78; 95% CI, 0.03–772.64; p =
0.55). The generalized estimating equation models found that both ln-renin
(adjusted odds ratio, 1.18; 95% CI, 1.02–1.37; p = 0.025) and
ln-lactate (adjusted odds ratio, 2.38; 95% CI, 1.05–5.37; p = 0.037)
were associated with mortality. Area under the receiver operating
characteristics curve analysis demonstrated that initial renin could predict
in-hospital mortality with fair discrimination (area under the receiver
operating characteristics curve, 0.682; 95% CI, 0.503–0.836; p =
0.05), but initial lactate could not (area under the receiver operating
characteristics curve, 0.615; 95% CI, 0.413–0.803; p = 0.27). Peak
renin (area under the receiver operating characteristics curve, 0.728; 95% CI,
0.547–0.888; p = 0.01) and peak lactate (area under the receiver
operating characteristics curve, 0.746; 95% CI, 0.584–0.876; p =
0.01) demonstrated moderate discrimination. There was no significant difference
in discriminative ability between initial or peak renin and lactate
concentration. At each study time point, a higher proportion of renin values
exceeded the threshold of normal (40 pg/mL) in nonsurvivors than in survivors,
but this association was not significant for lactate.
Conclusions:
Although there was no significant difference in the
performance of renin and lactate when examining the absolute values of each
laboratory, a positive rate of change in renin concentration, but not lactate
concentration, over 72 hours was associated with in-hospital mortality. For
each one-unit increase in the slope of ln-renin, the odds of mortality
increased 10-fold. Renin levels greater than 40 pg/mL, but not lactate levels
greater than 2 mmol/L, were associated with in-hospital mortality. These
findings suggest that plasma renin kinetics may be superior to lactate kinetics
in predicting mortality of hypotensive, critically ill patients.
