Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Fluid overload before continuous hemofiltration and survival in critically ill children: a retrospective analysis.

Foland JA, Fortenberry JD, Warshaw BL, et al.

Crit Care Med. 2004 Aug;32(8):1771-6. [abstract]

Reviewed by Linda Thompson MD, Texas Children's Hospital, Baylor College of Medicine, Houston TX

Review posted January 18, 2005

I. What is being studied?:

The study objective:

The authors aimed to identify risk factors for mortality before CVVH (continuous venovenous hemofiltration) initiation in critically ill children. They hypothesized that patient survival would be associated with degree of fluid overload before initiation of CVVH, with severity of illness, and with primary clinical diagnosis.

The study design:

Retrospective cohort study

The patients included:

113 patients who had received CVVH in either the CVICU, PICU or NICU at Children's Healthcare of Atlanta at Egleston (a tertiary referral center) over a 5 year period (November, 1997 to January, 2003). Patients were identified using the institutional CVVH database.

The patients excluded:

Patients receiving ultrafiltration without ultrafiltration replacement or those receiving CVVH/ultrafiltration while also receiving ECMO.

The predictive indices compared:

1. Age
2. Weight
3. PRISM III at time of CVVH initiation
4. Hospital days before CVVH initiation
5. Days in ICU before CVVH initiation
6. Estimated GFR at CVVH initiation
7. P/F ratio (P02 / Fi02) at CVVH
8. % Fluid overload (= total fluid intake - total fluid output [L] / body weight [kg] x 100) up to 7 days before CVVH initiation
9. Intubated
10. Requiring vasoactive infusions
11. Body surface area
12. Duration of CVVH
13. Ultrafiltration rates
14. Serum creatinine

The outcomes assessed:

Survival to ICU discharge

II. Are the results in the study valid?

Primary questions:

1. Was there a representative and well-defined sample of patients at a similar point in the course of the disease?

The patient sample was drawn from patients with both primary and secondary renal disease in several areas of the hospital and is representative of children receiving CVVH in most tertiary institutions. Patients received CVVH primarily for acute fluid overload and acute renal failure (67%), but also for electrolyte imbalance (25%), acute on chronic renal failure (6%) and hyperammonemia (2%). PRISM III scores ranged from less than 9 to greater than 17, but it is difficult to judge whether patients were at a similar point in the course of the disease simply based on PRISM scores given the variety of indications for CVVH and the absence of defined criteria for initiating CVVH.

2. Was follow-up sufficiently long and complete?

There is likely a small chance that patients who survived to ICU discharge may have gone on to die later in their hospital course.

Secondary questions:

3. Were objective and unbiased outcome criteria used?

Survival to ICU discharge is an objective and easily measured outcome criterion, though the time of ICU discharge may vary from institution to institution according to capabilities of step-down units.

4. Was there adjustment for important prognostic factors?

Yes, multivariable analysis was done in all children, children with MODS (Multiple Organ Dysfunction Syndrome as defined by Wilkinson et al (1) and Proulx et al (2)), and children with ≥ 3 organ MODS.

III. What are the results?

1. How large is the likelihood of the outcome event(s) in a specified period of time?

Overall, survival was 61% (69/113). Survival was 90% in patients with primary renal disease. Survival was 51% in patients with secondary renal disease (with a range of 19% survival in BMT patients (3/16) to 100% survival in patients with respiratory failure (7/7)). For patients with ≥ 3 organ MODS, survival was 52%. Survivors had lower PRISM III scores (median 12 vs. 15.5, p < 0.001), fewer hospital days before initiating CVVH (median 3 vs. 15.5, p < 0.001), fewer days in ICU before initiating CVVH (median 1 vs. 3, p < 0.03), lower % FO (median 7.8% vs. 15.1%, p = 0.02), were less likely to be intubated (median 58% vs. 91%, p < 0.001), and less likely to require vasoactive infusions (median 61% vs. 84%, p = 0.009). Weight, age, duration of CVVH, GFR and ultrafiltration rates were not significantly different between the two groups.

In children with MODS, survivors had lower PRISM III scores (median 12 vs. 15.5, p = 0.002), fewer hospital days before initiating CVVH (median 3 vs. 15.5, p = 0.001), and lower % FO (median 9.3% vs. 15.1%, p = 0.08). Though not statistically significant, children with MODS who survived had fewer ICU days before initiating CVVH (median 2 vs. 3, p = 0.12) and a lower percentage receiving vasoactive infusions (median 71.2% vs. 84.1%, p = 0.13).

In children with ≥ 3 organ MODS, survivors had lower PRISM III (median 14 vs. 16, p = 0.02), fewer hospital days before initiating CVVH (median 3.5 vs. 16, p = 0.001), and lower % FO (median 9.2% vs. 15.5%, p = 0.01).

As shown in the table below, by multivariable logistic regression model using stepwise selection, an increase of PRISM III score by 2 points or a 10% increase in fluid overload show a trend towards increasing the odds of death in all groups. 10% fluid overload may increase the odds of death up to 2 - 3 times that of children without fluid overload.

All children receiving CVVH	OR	95% CI	p value
PRISM III (per 2 point increase)	1.24	1.02 - 1.5	0.03
Fluid overload (per 10% increase)	1.37	0.97 - 1.94	0.07
Children with MODS
PRISM III (per 2 point increase)	1.21	0.99 - 1.47	0.06
Fluid overload (per 10% increase)	1.30	0.92 - 1.84	0.13
Children with ≥ 3 organ MODS
PRISM III (per 2 point increase)	1.10	0.88 - 1.39	0.40
Fluid overload (per 10% increase)	1.78	1.13 - 2.82	0.01

2. How precise are the estimates of likelihood?

Confidence intervals include an odds ratio of 1 or very close to 1 in most groups so the odds ratios must be interpreted with caution. The most significant finding appears to be the increased odds of death with 10% increase in fluid overload in children with ≥ 3 organ MODS. The 95% CI reveals an increased risk in this group, up to an OR of 2.82, and the p value shows an independent association with mortality. While the PRISM III score loses discrimination in this group, the percent fluid overload remains an independent predictor of mortality.

IV. Will the results help me in caring for my patients?

1. Were the study patients and their management similar to my own?

Yes, the age range and reasons for initiation of CVVH are similar to that seen in most tertiary institutions.

2. Will the results lead directly to selecting or avoiding therapy?

The study provides indirect evidence for considering CVVH before significant fluid overload, especially in ≥ 3 organ MODS. The retrospective nature of this study, and in particular, the non-prococolized decision-making in initiating CVVH, limits the strength of this conclusion.

3. Are the results useful for reassuring or counseling patients?

While it was not the authors' intent to develop a prediction tool, the study's large sample of pediatric patients adds strength to previous studies showing evidence of increased mortality with multi-organ dysfunction. Certainly patients cannot be reassured that mortality risk will be decreased with CVVH.

References:

1. Wilkinson JD, Pollack MM, Ruttimann UE, et al: Outcome of pediatric patients with multiple organ system failure. Crit Care Med 1986; 14:271Ð274 [abstract]
2. Proulx F, Gauthier M, Nadeau D, et al: Timing and predictors of death in pediatric patients with multiple organ system failure. Crit Care Med 1994; 22:1025Ð1031 [abstract]

 

Comments

Back to the EB Journal Club Index