FRCA Notes

Sepsis is: 'a life-threatening organ dysfunction caused by a dysregulated host response to infection'

Septic shock is: 'sepsis with circulatory and cellular/metabolic abnormalities profound enough to substantially increase mortality'
I.e. it is a subset of septic patients, who:

Despite adequate volume resuscitation
Have a lactate >2mmol/L
And a vasopressor requirement to keep MAP >65mmHg

Septic shock is associated with an in-hospital mortality of 40% of more

Early identification and appropriate management in the initial hours after the development of sepsis can improve outcome
Organ dysfunction in sepsis can be represented by an increased Sequential Organ Failure Assessment (SOFA) score of ≥2
A SOFA score of ≥2 is associated with an in-hospital mortality of >10%

For patients in non-ICU settings, patients with infection can be identified as having poorer outcomes from sepsis if they fulfil two of the quick-SOFA (qSOFA) critera:

RR >22/min
SBP <100mmHg
Altered mentation

The APACHE II score is general ICU mortality predictor tool, for use at time of admission to ICU

Cultures should be taken prior to initiating antibiotic therapy (FABLED study)
One should not delay administration of antibiotics however, with an oft-quoted 7.6% increased mortality for each hour delay
Give broad-spectrum antibiotics targeting the likely pathogen, altered according to known sensitivites
The results of the BLISS and BLING-II trials suggest no mortality difference between continuous or intermittent bolus administration of antibiotics; the results of BLING-III are awaited

One should consider non-bacterial causes of sepsis, particularly in those with risk factors e.g. immunosuppression
Using β-D-glucan to guide early anti-fungal treatment didn't improve 28-day mortality in the CandiSep trial (2022)

Fluid choice

Surviving Sepsis Guidelines suggest crystalloid
Balanced crystalloid is probably better than 0.9% NaCl

The SAFE and ALBIOS studies demonstrated albumin solutions are probably not more harmful than crystalloid, but equally aren't more effective
The CHEST and 6S studies showed hydroxyethyl starch boxes kidneys and shouldn't be used

The TRISS study showed that a transfusion threshold of 70g/L was as effective as 90g/L

Fluid volume

Surviving Sepsis Guidelines suggest at least 30ml/kg (IBW) of crystalloid within three hours
This feels like a lot of fluid, and there's some suggestion excess fluid is bad:

SOAP study: positive fluid balance among the strongest prognostic factors for death
VASST study: positive fluid balance an independently associated with mortality
CLASSIC study: no difference in 90-day mortality between standard and restricted fluid regimens
CLOVERS study: no difference in 90-day mortality between standard and restricted fluid regimens
No 90-day mortality outcome difference between restrictive and liberal fluid strategies (NEJM, 2023)

Over at the kid's table:

The (pilot) FiSh trial found difference in outcome between 10ml/kg and 20ml/kg boluses for septic shock
The FEAST study found fluid boluses worsen 48hr mortality, but the external validity of this African study is low

Fluid responsiveness

Surviving Sepsis Guidelines suggest use of dynamic measures such as lactate and CRT to guide fluid resuscitation
There's no one good measure:

Capillary refill time the best of a bad bunch, in that it actually correlates with responsiveness, but the results can be affected by other pathology
There are caveats associated with most other markers such as CVP, PAWP, SVV, PPV, GEDVI, and IVC diameter
Passive leg raise autotransfusion has caveats attached to it too, but is probably not too bad

When to stop

Early goal-directed therapy (EGDT), i.e. a protocolised fluid resuscitation programme, was historically in vogue but has fallen away owing to the results of several trials:

ProCESS: No significnat morbidity or mortality advantage of protocol-based resuscitation
ARISE: EGDT did not improve 90-day all-cause mortality
ProMISe: No difference in 90-day survival between standard care and strict EGDT

One can still use a bunch of measurements to help assess when a patient may be fluid replete:

Suggestion the patient is no-longer fluid responsive:

PPV <12%
SVV <10%

Haemodynamic targets achieved:

MAP >65mmHg
CVP >8cmH₂O

Suggestion there is adequate end-organ perfusion:

CRT <3s
Lactate <2mmol/L
Urine output >0.5ml/kg/hr
S_cvO₂ >70%
Resolving clinical features of hypovolaemia

What MAP to target?

Surviving sepsis guidelines suggest a MAP target of 65mmHg

SEPSISPAM showed targeting a higher (80 - 85mmHg) MAP, when compared to lower (65 - 70mmHg), led to:

Higher rates of AF in the higher MAP target group
Lower rates of AKI and RRT in the higher MAP target group if the patient had chronic hypertension

The 65-Trial found no difference in mortality between a MAP target of 60 - 65mmHg vs. standard care in those >65yrs

Which drug to use first?

Noradrenaline is typically used as the first line agent, as it is:

Non-inferior to vasopressin (VASST study)
Non-inferior to adrenaline, but with fewer side-effects
Superior to dopamine

There's some evidence earlier use of a vasopressor is better for outcome

What next?

If evidence of low cardiac output, add a positive inotrope e.g. dopamine

Once noradrenaline is >0.2μg/kg/min add in vasopressin

Benefits from a catecholamine-sparing effect and an endogenous ADH-like (fluid-retention) effect
Suffers from causing splanchnic and coronary vasoconstriction, as well as platelet aggregation

Add in hydrocortisone (200mg/24hrs) for putative, relative, adrenal insufficiency

ADRENAL study: no difference in 90-day mortality vs. placebo in septic shock
HYPRESS study: use of hydrocortisone does not reduce the development of septic shock in those with sepsis
CORTICUS study: no survival benefit in septic shock, but does appear to reverse shock quicker than placebo
Some suggestion one should only give it if noradrenaline is >0.5μg/kg/min, and should only continue if the noradrenaline dose has decreased by 50% in 24hrs

Pathophysiological effects

Sepsis is characterised by widespread release of cytokines, ROS and proteases leading to both direct and indirect cell damage
Ensuing vasodilation and capillary leak causing relative and absolute hypovolaemia
Resuscitation may subsequently lead to considerably increased total body water

Microcirculatory blood flow is impaired, causing:

Heterogenous organ perfusion
Mitochondrial dysfunction
Cellular hypoxia
Subsequent organ dysfunction and failure

The extent of pharmacokinetic changes will be dynamic and highly variable owing to the interplay of:

Infectious factors such as causative organisms
Patient factors such as age, physiology, comorbidities
Treatment factors such as fluid resuscitation, vasoactive drugs

Absorption

Absorption is generally reduced, owing to:

Critical illness
Use of opioids
GI hypoperfusion from shock and/or venous congestion from fluid resuscitation

Use of vasoactive drugs in sepsis further decreases splanchnic blood flow and gut perfusion
This may impair enteral absorption of certain regular medications including antiretroviral, anti-Parkinsonian and certain psychoactive drugs

Redistributed blood flow in shock causes decreased skin and muscle perfusion
This leads to unpredictable absorption of subcutaneous drugs such as LMWH and insulin

Distribution

The redistribution of blood away from peripheral tissues can decrease> the volume of distribution of some fat-soluble medications
This increases plasma concentration and may lead to adverse effects

V_D can also be decreased by acidaemia, leading to increased ionisation of weak bases such as opioids and local anaesthetics

Conversely endothelial damage, capillary 'leak' and fluid resuscitation may significantly increase the V_D of hydrophilic medications
This reduces plasma concentrations and potential underdosing; examples include β-lactam and aminoglycoside antibiotics
Existing comorbidities (cardiac and liver failure) or extracorporeal circuits can further increase V_D

Low serum albumin concentrations initially lead to an increase in the free fraction of highly protein-bound drugs
This is counteracted by increased volume of distribution and increased clearance
As such, highly protein-bound drugs may require increasing loading/maintenance doses
By contrast, midazolam has a more rapid onset in the presence of hypoalbuminaemia

Metabolism

The majority of drugs undergo hepatic metabolism
In sepsis, hepatic (enzyme) dysfunction arises from:

Hypoxic hepatitis
Sepsis-induced cholestasis
Pro-inflammatory cytokines directly impairing CYP450 function
Reno-hepatic cross-talk, further impairing CYP450 function through cytokine & non-cytokine-mediated mechanisms
Therapeutic hypothermia
Drug interactions e.g. with PPI's, macrolides, fluoroquinolones or azole antifungals

Hepatic blood flow is also reduced, by factors such as vasoactive drugs, PPV and prone positioning

The net effect of these pathophysiological changes is:

An impairment of drug delivery to hepatocytes
Impaired hepatocyte ability to extract the oxygen required for drug metabolism
Reduced clearance of drugs with high hepatic extraction ratios, such as propofol and fentanyl

Elimination

AKI is common in sepsis and will impair elimination of drugs relying on renal excretion
This may lead to an accumulation of certain drugs, including glycopeptides, aminoglycosides and β-lactams
Elimination may be further altered by the use of RRT

Sepsis