Thread by @Dr_Hari_Krishna

Thread re: evidence appraisal/paucity of #PedsICU data based on recent @ESPNIC_Society guidelines on IV maintenance fluids
link.springer.com/content/pdf/10… @yourICM

Started by discussions in our monthly #PICUJournalWatch sessions @Bham_Childrens #PedsICU based on @PICJournalWatch
1/

PICO3: The paper states [& I don't disagree]:
In "critically ill children", balanced solutions should be favoured when prescribing *intravenous maintenance* fluid therapy to slightly reduce length of stay:
B Strong consensus
& "in acutely ill children": A Strong consensus 2/

They add "balanced IV-MFT solutions, the length
of acute care or PICU stay were slightly but significantly decreased in children receiving balanced solutions in a meta-analysis of 5 studies, including 283 patients. mean difference: −0.20 days; 95% CI [−0.33; −0.08], p=0.001 3/

Paper states "Implications for practice: Third, balanced fluids should be the standard IV-MFT solution used in children."

Let's take a plunge using #MetaAnalysis as the example:

4/

Paper #1: Balamuth 2019: PROMPT bolus study of fluid resuscitation in sepsis NS v RL. Same fluids were also used for maintenance, but resuscitation was the focus, I think. Recruited in ED. I doubt that the maintenance fluids [~50% of fluids infused] contributed to outcomes

5/

Paper#2: Kartha 2017: RL v NS RCT of Diarrhoea + dehydration. *Not "Maintenance*
They state "0.45% NS 5%dextrose ... was administered as maintenance fluid and for on-going losses (10 mL/kg ...) until the child was accepting ORS well. WHO ORS and oral feeds were started ..."
6/

Paper #3: Mahajan 2012: diarrhoea, dehydration, mainly Cholera-rapid IV rehydration RCT of RL v NS, *not maintenance*. They say "used either reduced osmolarity WHO ORS or...0.45% saline in 5% dextrose...as replacement fluids depending upon the child’s ability to drink"
7/

Paper 4 & 5 both were in kids with dehydration with DKA. Again, as dehydration correction. PICU/HDU LoS used in meta-analysis from Yung et al. (Hartmann's v NS), and Hospital LoS used from the SPINK trial (Plasmalyte v NS).
8/

Essentially ALL studies in meta-analysis were of fluid resuscitation and not maintenance therapy. Fig in tweet #3 of this thread will show heavy weight for Yung et al [That data however, is not hospital LoS unlike other studies!]
9/

Reproduce results in #RStats [packages:meta/dmetar-Github:Mathias Harrer] Using "mean difference" [MD] as reported by study: I get identical results with a statistically significant reduction in (some) LoS. Mean reduction of 0.2 days-But, clinical significance, unsure?

10/

Use #SMD instead of #MD in analysis:
No clinical or statistical significance. Effect disappears! Different distribution of weights & the study reporting only PICU/HDU LoS has lower weight here: *Highlights issues with paucity of data again, I think* @VallaFred @kalapappaj
11/

Authors rightly point out: "The main limitation of these evidence-based recommendations is the general paucity of evidence and the low level of evidence around some of the questions" & also:

12/

In my opinion, none of the studies included in meta-analysis of PICO3 were actually related to the impact of IV maintenance fluids. They were all used for fluid resuscitation--which is not what this guideline is about. *It's sad that there is paucity of #PedsICU literature *
13/

There is no point in just stating
"there was insufficient evidence to recommend x over y". The recommendations appear prudent.
I agree with those -- however, I am unsure PICO3 as an example, is based on correct data or correct outcomes. @Martin_Kneyber @lyvonnetume
14/

For clarity:
The PICO3 recommendation is probably correct.
**Lack of data is the issue I am highlighting**
Solution? I don't know
Publication of more #PedsICU research-How? Motivate a generation of #PedsICU #PedsCICU to engage w research, understand evidence-How?
15/15 END

.@threadreaderapp unroll please @PICJournalWatch