现在的位置: 首页研究点评, 进展交流>正文
[Blue Journal]: 渥太华大学推荐阅读(3/3)
2017年06月03日 研究点评, 进展交流 暂无评论

Recommended Reading from the University of Ottawa Nephrology Fellows

David Massicotte-Azarniouch, Syed Obaid Amin, Caitlin Hesketh and Edward G. Clark

AJRCCM Articles in Press. Published on 02-May-2017 as 10.1164/rccm.201611-2375RR

Bitker L, et al.; Prevalence and risk factors of hypotension associated with preload-dependence during intermittent hemodialysis in critically ill patients. Crit Care (10)

Reviewed by Caitlin Hesketh

Intradialytic hypotension (IDH) is a common complication of intermittent hemodialysis (HD) in the ICU and may be associated with increased mortality and a lower likelihood of renal recovery after dialysis-requiring AKI (11). While IDH is frequently presumed to be secondary to preload reduction from ultrafiltration, multiple other RRT- and patientrelated factors may be responsible to varying degrees (12).

Bitker et al. (10) performed a prospective, observational, single-center study to assess the prevalence of, and risk factors for, IDH in critically ill patients receiving intermittent HD in ICU. At the time of their first occurrence, IDH episodes (defined as MAP < 65 mmHg) were classified as either related or unrelated to cardiac preload according to changes in cardiac output following a passive leg raise (PLR) (3). The secondary objectives were to identify risk factors for IDH and to assess the diagnostic performance of variables associated with preload-dependent hypotension.

Forty-seven patients met inclusion criteria, one of which was ongoing use of a PiCCO® device, previously described and validated for real-time cardiac output monitoring (13). The median SOFA score was 8 with 36% of patients receiving mechanical ventilation and 50% on vasopressors at the time of the intermittent HD sessions’ onset. Preload dependence was defined as ≥10% increase in stroke volume within 1 minute of PLR. Of 107 HD sessions, 61 (57%, 95% CI: 47-66%) were complicated by IDH after a median of 35 minutes. Only 11 of 61 hypotensive episodes (19%, 95% CI: 10-31%) were defined as preload dependent. Risk factors for hypotension related to preload dependence were pulmonary vascular permeability index (PVPI) (a measure of the permeability of the pulmonary alveo-capillary barrier calculated as the ratio of extravascular lung water to pulmonary blood volume (14)) ≥1.6 (sensitivity 91%, specificity 53%) and mechanical ventilation. SOFA scores were significantly higher and MAP was significantly lower at the onset of sessions complicated by IDH compared with sessions that were not. There was no significant difference in the prevalence of preload dependence at session outset, vasopressor dose, or hemodynamic measurements (other than MAP) between those sessions complicated by IDH and those that were not.

In this study, very few patients were preload dependent prior to intermittent HD sessions. The surprising finding was that less than 20% of sessions complicated by IDH were associated with preload-dependence at the time IDH first occurred. This and the relatively early occurrence of IDH, prior to much fluid removal having occurred, suggests IDH may not be primarily due to excessive ultrafiltration in many patients. Recent studies in the end-stage kidney disease HD population suggest that HD-related myocardial stunning, unrelated to fluid removal or underlying cardiovascular disease, is frequently an important culprit (15). Multiple studies have shown an association between fluid overload and worse outcomes in critically ill patients (16). This raises the further question as to whether, for some critically ill and volume overloaded patients, other methods of hemodynamic and cardiovascular optimization should be preferred over empiric discontinuation of fluid removal in response to IDH (e.g. would it be beneficial to aggressively increase vasopressor doses in order to achieve ordered fluid removal targets rather than reduce those targets in response to IDH as is typically done).

Limitations of this study include the very small number of patients included from a singlecenter and its observational design. As well, PiCCO® measurements following PLR may have been unreliable in the setting of acute hypotension and PLR may be less useful in effecting a change in preload when performed from the supine position, as was the case with patients in this study. The overall small number of patients with preload dependent IDH prevented a multivariate analysis that might have better defined the relevance of various factors with which it was associated. Certainly, myriad other RRT- and patientrelated factors might predispose patients to experience IDH (e.g. low calcium concentration dialysate, septic shock). Nonetheless, this study sheds new light on IDH mechanisms in the critically-ill by suggesting that much of the IDH observed in this population occurs independently of fluid removal. As such, its intriguing results warrant further study.

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博