Braun Chloe, Takeuchi Tomonori, Lambert Josh, Liu Lucas, Roberts Sarah, Carter Stuart, Beaubien-Souligny William, Tolwani Ashita, Neyra Javier A
The University of Alabama at Birmingham, Birmingham, AL, USA.
Tokyo Medical and Dental University, Tokyo, Japan.
J Intensive Care. 2024 Dec 31;12(1):55. doi: 10.1186/s40560-024-00772-w.
Fluid balance gap (FBgap-prescribed vs. achieved) is associated with hospital mortality. Downtime is an important quality indicator for the delivery of continuous renal replacement therapy (CRRT). We examined the association of CRRT downtime with FBgap and clinical outcomes including mortality.
This is a retrospective cohort study of critically ill adults receiving CRRT utilizing both electronic health records (EHR) and CRRT machine data. FBgap was calculated as achieved minus prescribed fluid balance. Downtime, or percent treatment time loss (%TTL), was defined as CRRT downtime in relation to the total CRRT time. Data collection stopped upon transition to intermittent hemodialysis when applicable. Linear and logistic regression models were used to analyze the association of %TTL with FBgap and hospital mortality, respectively. Covariates included demographics, Sequential Organ Failure Assessment (SOFA) score at CRRT initiation, use of organ support devices, and the interaction between %TTL and machine alarms.
We included 3630 CRRT patient-days from 500 patients with a median age of 59.5 years (IQR 50-67). Patients had a median SOFA score at CRRT initiation of 13 (IQR 10-16). Median %TTL was 8.1% (IQR 4.3-12.5) and median FBgap was 17.4 mL/kg/day (IQR 8.2-30.4). In adjusted models, there was a significant positive relationship between FBgap and %TTL only in the subgroup with higher alarm frequency (6 + alarms per CRRT-day) (β = 0.87 per 1% increase, 95%CI 0.48-1.26). No association was found in the subgroups with lower alarm frequency (0-2 and 3-5 alarms). There was no statistical evidence for an association between %TTL and hospital mortality in the adjusted model with the interaction term of alarm frequency.
In critically ill adult patients undergoing CRRT, %TTL was associated with FBgap only in the subgroup with higher alarm frequency, but not in the other subgroups with lower alarms. No association between %TTL and mortality was observed. More frequent alarms, possibly indicating unexpected downtime, may suggest compromised CRRT delivery and could negatively impact FBgap.
液体平衡差距（规定值与实际值之间的差距）与医院死亡率相关。停机时间是持续肾脏替代治疗（CRRT）实施的一项重要质量指标。我们研究了CRRT停机时间与液体平衡差距及包括死亡率在内的临床结局之间的关联。
这是一项对接受CRRT的危重症成年患者的回顾性队列研究，利用电子健康记录（EHR）和CRRT机器数据。液体平衡差距的计算方法为实际液体平衡减去规定液体平衡。停机时间，即治疗时间损失百分比（%TTL），定义为CRRT停机时间与总CRRT时间的比值。在适用的情况下，当患者转为间歇性血液透析时停止数据收集。分别使用线性回归模型和逻辑回归模型分析%TTL与液体平衡差距及医院死亡率之间的关联。协变量包括人口统计学数据、CRRT开始时的序贯器官衰竭评估（SOFA）评分、器官支持设备的使用情况，以及%TTL与机器警报之间的相互作用。
我们纳入了500例患者的3630个CRRT患者日，患者中位年龄为59.5岁（四分位间距50 - 67岁）。患者CRRT开始时的SOFA评分中位数为13（四分位间距10 - 16）。%TTL中位数为8.1%（四分位间距4.3 - 12.5），液体平衡差距中位数为17.4 mL/kg/天（四分位间距8.2 - 30.4）。在调整模型中，仅在警报频率较高的亚组（每CRRT日警报6次及以上）中，液体平衡差距与%TTL之间存在显著正相关（每增加1%，β = 0.87，95%置信区间0.48 - 1.26）。在警报频率较低的亚组（0 - 2次和3 - 5次警报）中未发现关联。在包含警报频率交互项的调整模型中，没有统计学证据表明%TTL与医院死亡率之间存在关联。
在接受CRRT的危重症成年患者中，%TTL仅在警报频率较高的亚组中与液体平衡差距相关，而在警报频率较低的其他亚组中则无关联。未观察到%TTL与死亡率之间存在关联。更频繁的警报可能表明意外停机，这可能提示CRRT实施存在问题，并可能对液体平衡差距产生负面影响。
