Dr. Hallie Prescott's presentation addresses the importance of de-escalating antibiotic treatment and managing fluid overload in patients with sepsis and critical illness. Evidence suggests that shorter courses of antibiotics are safe and can effectively limit the development of resistance, while careful management of fluid balance is vital to minimize complications like new disability. The proposed approach includes stopping antibiotics based on specific clinical criteria and emphasizing the restoration of euvolemia after stabilization.

1. Hallie Prescott, MD, MSc
The Back-End of Sepsis:
De-escalating & de-resuscitating
@HalliePrescott
B17 - CRITICAL CARE:
BURDEN OF SURVIVAL -
OUTCOMES AFTER
CRITICAL ILLNESS IN
ADULTS AND CHILDREN
*ThispresentationdoesnotrepresentviewsofUSGovernmentorDepartmentofVeteransAffairs.

2. Rhodes, et al. Crit Care Med. 2017.

3. Fleischmann, et al. AJRCCM. 2016.
19.4 million cases 5.3 million deaths
An estimated 14.1 million patients survive
(severe) sepsis each year.

4. microbiome
disruption
edema,
fluid overload
recurrent
sepsis
disability

5. "Make sure to finish
your course, or you
might breed resistance!”
When I was in training… this was easy!
“For complicated UTI,
treat for 7-14 days”
7 days if low-risk
10 days if medium-risk
14 days if high-risk

6. Royer, et al. J Hosp Med, 2018.
19 Studies
***
2867 Patients
Difference: 1.6% (-1.0%, 4.2%). Difference: 0.3% (-1.2%, 1.8%).

7. Royer, et al. J Hosp Med, 2018.
Conclusions:
• Shorter courses are safe in common infections
• Limited data on longer-term risks (e.g. secondary infection,
emergence of MDR pathogens)

8. “If you use penicillin,
use enough!”
-Alexander Fleming,
1945 Nobel Prize Acceptance
Llewelyn, et al. BMJ, 2017.
HIV, Tuberculosis
Target Selection:
Microbes develop resistance
during treatment
Collateral Selection:
Antibiotics select for normal flora and
pathogens with resistance genes
Staph, Enterobacter, Klebsiella

9. Patients: 312 patients hospitalized with pneumonia
Intervention: Symptom-guided protocol
(afebrile x 48 hours, HR < 100, RR < 24, SpO2 > 90)
Control: Usual care
Outcomes: Clinical success, symptom scores at 10, 30d
Uranga, et al. JAMA IM, 2016.

10. Uranga, et al. JAMA IM, 2016.
10
5
0
2
4
6
8
10
12
1 2
Shorter antibiotic duration No difference in:
clinical resolution
symptoms scores
p<0.001

11. Patients: 1575 ICU patients with suspected bacterial infection
Intervention: Procalcitonin-guided antibiotic cessation
Control: Usual care
Outcomes: Mortality, Antibiotic exposure
De Jong, et al. Lancet Infect Dis. 2016.

12. De Jong, et al. Lancet Infect Dis. 2016.
9.3 7.5
0
1
2
3
4
5
6
7
8
9
10
1 2
Shorter antibiotic duration
27% 20%
0%
5%
10%
15%
20%
25%
30%
1 2
Lower hospital mortality
p<0.001 p=0.02

13. Iankova, et al. Crit Care Med. 2018. Lam, et al. Crit Care Med. 2018.
Wirtz, et al. Crit Care. 2018.

14. Iankova, et al.
CCMed, 2018
Lam, et al.
CCMed, 2018
Wirtz, et al.
Crit Care, 2018
Population
Confirmed or
suspected sepsis
ICU patients
ICU patients
with systemic infection
Studies 10 9 Cessation
11
*individual patient data
Patients 3,489 3,521 4,482
Antibiotic
Duration
Mortality
Iankova, et al. Crit Care Med. 2018. Lam, et al. Crit Care Med. 2018.
Wirtz, et al. Crit Care. 2018.

15. Iankova, et al.
CCMed, 2018
Lam, et al.
CCMed, 2018
Wirtz, et al.
Crit Care, 2018
Population
Confirmed or
suspected sepsis
ICU patients
ICU patients
with systemic infection
Studies 10 9 Cessation
11
*individual patient data
Patients 3,489 3,521 4,482
Antibiotic
Duration
-1.5 days -1.3 days -1.2 days
Mortality
Greater in less
sick patients
Iankova, et al. Crit Care Med. 2018. Lam, et al. Crit Care Med. 2018.
Wirtz, et al. Crit Care. 2018.

16. Iankova, et al.
CCMed, 2018
Lam, et al.
CCMed, 2018
Wirtz, et al.
Crit Care, 2018
Population
Confirmed or
suspected sepsis
ICU patients
ICU patients
with systemic infection
Studies 10 9 Cessation
11
*individual patient data
Patients 3,489 3,521 4,482
Antibiotic
Duration
-1.5 days -1.3 days -1.2 days
Mortality RR 0.9 (p=0.11)
RR 0.87
(p=0.02)
OR 0.89 (p=0.03)
Greater in less
sick patients
Equal across all
subgroups
Iankova, et al. Crit Care Med. 2018. Lam, et al. Crit Care Med. 2018.
Wirtz, et al. Crit Care. 2018.

17. In summary:
1. Duration recommendations based on fear of resistance and
treatment failure.
2. Existing data suggests shorter courses sufficient.
3. Treatment response and/or biomarker trajectory can be used to
safely limit antibiotics.

18. My current approach to antibiotic de-escalation:
Stop antibiotics once any 1 of criteria is met*:
• You decide the patient probably wasn’t infected
• The patient is clinically improved
• The patient is still sick, but procalcitonin is normalized
• The patient has received 7-8 days of Abx
(*Caveat: unless there are specific conditions warranting longer treatment,
e.g. endocarditis, staph bacteremia, TB, etc.)

19. Mitchell, et al. AnnalsATS. 2015.
Volume overload  2x higher odds new disability, facility discharge
positive fluid balance
86%
36%
volume overload
42%
ever got diuretic

20. De-escalating fluids
Vincent and de Backer. Circulatory Shock. NEJM. 2013.
Finfer, Myburgh, Bellomo. Intravenous fluid therapy in critically ill adults. Nat Rev Nephology. 2018.

21. My current approach to de-resuscitation
• Diuresis once hemodynamically stable and off vasopressors x 24 hrs
• Target euvolemic exam / “dry weight”
• Recognize that muscle mass loss is common

22. Patient Case
Pre-hospital
120 kg
ICU admit
115 kg
ICU Day 7
130 kg
ICU Day 16
120 kg
ICU Day 18
112 kg

23. Thank
You!
@hallieprescott
Stop antibiotics once any 1 of criteria is
met*:
• You decide the patient probably wasn’t
infected
• The patient is clinically improved
• The patient is still sick, but procalcitonin
is normalized
• The patient has received 7-8 days of Abx
(*Caveat: unless there are specific conditions warranting
longer treatment, e.g. endocarditis, staph bacteremia, TB,
etc.)
Target “Dry Weight”
• Diuresis once hemodynamically
stable and off vasopressors x 24 hrs
• Recognize that muscle mass loss is
common