Appropriate timing and dosing of

antibiotics in sepsis
Diana L. Wells, PharmD, BCPS
Assistant Clinical Professor
Auburn University Harrison School of Pharmacy
Auburn, Alabama

Jeffrey Fish, PharmD, BCPS

Clinical Pharmacist, Trauma and Life Center
University of Wisconsin Hospital and Clinics
Madison, Wisconsin
 Objectives
1. Summarize literature supporting appropriate
choice and timing of antibiotics in sepsis
2. Using a patient case, develop an
antimicrobial dosing regimen to achieve
early and optimal exposure to appropriate
antimicrobial agents
3. Recognize patient factors which may impact
antibiotic dosing for septic patients
 Outline – Part 1: Timing of
antibiotics in sepsis
• Guideline recommendations
• Literature supporting early, appropriate
antibiotics
• Example antibiotic regimens for sepsis
• Overcoming barriers to timely antibiotic
administration
 Guideline recommendations
 Administration of effective IV antimicrobials
within the 1st hour of recognition of septic shock
(grade 1B) and severe sepsis without septic
shock (grade 1C)
 Initial empiric anti-infective therapy of one or
more drugs that have activity against all likely
pathogens and that penetrate in adequate
concentrations into tissues presumed to be the
source of sepsis (grade 1B)

Crit Care Med 2013;41:580-637

 Early, appropriate antibiotics
 Early = within 1 hour after recognition
of potential septic shock

 Appropriate = in vitro activity against

pathogen
 Route of administration
 Dose and frequency
 Penetration
 Cidality

Crit Care Clin 2011;27:53-76

Fraction of total patients
Effect of timing on survival

Time from hypotension onset (hours)

Adapted with permission from:

Crit Care Med 2006;34:1589-96
 Effect of inappropriate antibiotics
on survival
Appropriate Inappropriate
OR (95% CI)
(n=4579) (n=1136)
Survived 52 10.3 9.45 (7.74 – 11.54)
P value
Immuno-
15 19.8 < 0.05
suppressed*
COPD 13.6 14.1 < 0.05
Dialysis 7.3 10.7 < 0.05
All numbers expressed as % unless otherwise specified
* Immunosuppression = chemotherapy or chronic steroids (>10mg prednisone daily)

Chest 2009;136:1237-48
 Risk Factors
MDR/Health-care associated
Fungemia
pathogens
• broad spectrum antibiotics within 90 d • broad-spectrum antibiotics
• hospitalization >5 d • central venous catheter
• local high antibiotic resistance rates • parenteral nutrition
• residence in LTCF • renal replacement therapy in ICU
• chronic dialysis within 30 d • neutropenia
• home wound care • hematologic malignancy
• family member with MDR infection • implantable prosthetic devices
• mechanical ventilation ≥5 d • immunosuppression
• immunosuppression • chemotherapy
• structural lung disease
• IV drug use
• COPD (Pseudomonas spp.)
• Influenza infection (MRSA)

Clin Infect Dis 2007;44:S27-72 Clin Infect Dis 2009;49:1-45

Am J Respir Crit Care Med 2005;171:388-416 Clin Infect Dis 2009;48:503-35
 Guideline recommendations
 Combination empirical therapy for the following
patients (grade 2B):
• Neutropenic with severe sepsis and for patients
with difficult-to-treat, multidrug-resistant bacterial
pathogens (Acinetobacter or Pseudomonas
bacteremia)
• Severe infections associated with respiratory
failure and septic shock (Pseudomonas
bacteremia)
• Septic shock from bacteremic Streptococcus
pneumoniae

Crit Care Med 2013;41:580-637

 Combination therapy vs.
monotherapy for septic shock
Mortality rate *
Monotherapy Combination Rx
HR (95% CI)
(n=1223) (n=1223)
28-Day, % 36.3 29 0.77 (0.67 – 0.88)
ICU, % 35.7 28.8 0.75 (0.63 – 0.88)
Hospital, % 47.8 37.4 0.69 (0.59 – 0.81)
# deaths
All Gram + , % 39.9 30.7 0.73 (0.58 – 0.92)
All Gram - , % 34.5 28.2 0.79 (0.67 – 0.94)
* Propensity score adjusted

Crit Care Med 2010;38:1773-85

 Antibiotic review: Sepsis from
pulmonary source
Infection Example antibiotic regimens
CAP β-lactam1 + azithromycin
β-lactam1 + respiratory FQ2
HCAP antipseudomonal β-lactam3
+ aminoglycoside4 or antipseudomonal FQ5
+ vancomycin or linezolid
1 ceftriaxone, cefotaxime, ampicillin/sulbactam
2 levofloxacin, moxifloxacin
3 piperacillin/tazobactam, cefepime, meropenem, imipenem, doripenem
4 gentamicin, tobramycin, amikacin
5 levofloxacin, ciprofloxacin

Clin Infect Dis 2007;44:S27-72

Am J Respir Crit Care Med 2005;171:388-416
Antibiotic review: Sepsis from catheter-
related bloodstream infection (CRBSI)
Infection Example antibiotic regimens
CRBSI vancomycin or daptomycin1
+ antipseudomonal β-lactam2,3
+/- aminoglycoside4
Fungemia + fluconazole or echinocandin5
risk factors
1 if high rates of vancomycin MIC ≥ 2 µg/mL
2 piperacillin/tazobactam, cefepime
3 meropenem, imipenem, doripenem
4 gentamicin, tobramycin, amikacin
5 caspofungin, micafungin, anidulafungin

Clin Infect Dis 2009;49:1-45

 Antibiotic review: Sepsis from
urinary source
Infection Example antibiotic regimens
Urosepsis 3rd generation cephalosporin1
+/- aminoglycoside2 or FQ3
Urological interventions or antipseudomonal β-lactam4,5
MDR risk factors
1 ceftriaxone, cefotaxime
2 gentamicin, tobramycin, amikacin
3 levofloxacin, ciprofloxacin
4 piperacillin/tazobactam, cefepime
5 meropenem, imipenem, doripenem

Int J Urol 2013; Epub ahead of print.

 Antibiotic review: Sepsis from
unknown source
Infection Example antibiotic regimens
Unknown antipseudomonal β-lactam1,2
+ aminoglycoside or antipseudomonal FQ3
+ vancomycin
Fungemia + fluconazole or echinocandin4
risk factors
1 piperacillin/tazobactam, cefepime
2 meropenem, imipenem, doripenem
3 levofloxacin, ciprofloxacin
4 caspofungin, micafungin, anidulafungin

Clin Infect Dis 2009;48:503-35

 Barriers to timely antibiotics
 Delayed recognition of sepsis and septic shock
• Infection
• Hypotension

 Inappropriate antimicrobial therapy

• Failure to use stat order
• Unrecognized risk factors for MDR pathogens
• No specifications for order of administration
• Logistical delays

Crit Care Clin 2011;27:53-76

 Impact of sepsis bundle
implementation
Achievement of bundle targets (n=15,022)
1st Quarter Final Quarter P value
Broad-spectrum
60.4 69.7 0.0002
antibiotics, %

 Administration of broad spectrum antibiotics

associated with lower hospital mortality
 OR (95% CI) = 0.86 (0.79–0.93)

Crit Care Med 2010;38:367-74

 Standardized order sets
Before (n=60) After (n=60) P value
Appropriate
71.7 86.7 0.043
antibiotics, %
28-day
48.3 30 0.04
mortality, %

Crit Care Med 2006;34:2707-13

 Overcoming barriers
 Education of healthcare professionals
• Multidisciplinary approach
• Medical Emergency Teams
 Update policies to minimize delays
• Administer antibiotics prior to transfer
• Order all initial IV antibiotics as stat
• Administer 1st dose of antibiotics as push
• Standardized treatment approach
• Symptom-based treatment pathway
• Sepsis protocols and order sets

Crit Care Clin 2011;27:53-76

Crit Care Med 2007;35:2568-75
 Take home points
 Evaluate risk factors for MDR/Health-care
associated pathogens
• Immunosuppression, COPD, hemodialysis,
LTCF residence

 Mortality reduction
• Combination antibiotics
• Sepsis bundles and protocols
• Early, appropriate antibiotics
Questions?
 Outline – Part 2: Dosing of
antibiotics in sepsis
• Pharmacokinetic differences in septic patients
• Antibiotic pharmacodynamic review
• Specific patient examples
 Pharmacokinetics
• Absorption
– Decreased gastric or subcutaneous absorption due to
shock and vasopressors
– Intravenous route preferred in severe sepsis / septic shock
• Oseltamivir
• Volume of distribution (Vd)
– Hydrophilic medications generally stay in the plasma
volume (Vd < 0.7 L/kg)
• Influenced by fluid administration and capillary leak
– Lipophilic medications distribute into intracellular and
adipose tissue (Vd > 1 L/kg)
• Not generally affected by fluid administration and third spacing
Crit Care Clin 2011;27:1-18
Crit Care Clin 2011;27:19-34
Crit Care Clin 2006;22:255-71
Chest 2012;141;1327-36
 Pharmacokinetics
• Metabolism
– Hepatic metabolism consists of two phases
• Phase 1: oxidation, reduction and hydrolysis
– Cytochrome P450
• Phase 2: glucuronidation, sulfation and acetylation
– Drugs can be classified by extraction ratio
• High (> 0.7): depends on hepatic drug flow
• Intermediate (0.3-0.7)
• Low (< 0.3): depends on hepatic (intrinsic) function
• Excretion
– Renal excretion is the primary excretory pathway for most parent
drugs or their metabolites
– Sepsis/shock patients frequently present with acute kidney injury
– May also present with increased renal excretion
• Augmented renal clearance
Crit Care Clin 2011;27:1-18
Crit Care Clin 2011;27:19-34
Crit Care Clin 2006;22:255-71
Chest 2012;141;1327-36
Pharmacodynamics

Clin Inf Dis 1998;26:1-12

Crit Care Clin 2011;27:1-18
Crit Care Clin 2011;27:19-34
Crit Care Med 2009;37:840-51
 Loading Doses
• Goal is to achieve therapeutic concentrations rapidly so loading
doses are usually recommended
• Recommend giving high end of normal loading dose (or even higher
dose)
– Example: Vancomycin (normal patient Vd ~0.7 L/kg)
• 100kg septic shock patient

• Recommended loading dose for complicated infections in seriously ill patients

is 25-30 mg/kg based on actual body weight
– Am J Health-Syst Pharm 2009;66:82-98
 Patient Case
• LL is a 45yo patient with a history of a renal
transplant in 2007 who presents with
respiratory distress and hypotension. He is
emergently intubated in the ER and fluid
resuscitated with 3L of NS.
– LL has NKDA, weighs 91kg and his admit SCr=3.2
mg/dl
• His SCr at a clinic visit one month prior = 1.3 mg/dl
– Cefepime, ciprofloxacin and vancomycin are
written for – What doses should be given?
 Renal Function – Acute Kidney Injury
• Lack of information in patients with sepsis/shock and acute kidney injury
– Since SCr is not at steady state -> not a reliable estimate of CrCl
– Concern for underdosing and treatment failure
• Recommendations from “A clinical update from Kidney Disease:
Improving Global Outcomes (KDIGO)”
– Loading dose: Volume of distribution is usually significantly increased in acute
kidney injury for hydrophilic medications
• Recommend: Aggressive loading doses (25-50% greater than normal)
– Maintenance dose: Need to estimate degree and rate of change in kidney status
• Need to also take into account nonrenal clearance
• Recommend: Initiate at normal or near-normal dosage regimens
– Therapeutic drug monitoring: Most concern for drugs with narrow therapeutic
window
• Recommend: Check serum concentrations if possible
• Recommend: If no serum concentrations: watch for excessive pharmacologic effect
or toxicity
– Concern with cefepime use in renal dysfunction (Hosp Pharm 2009;44:557-61)
• What dose to give?

Kidney International 2011;80:1122-37

 Revised Patient Case
• LL is a 45yo patient with a history of ESRD
(IHD Mon/Wed/Fri) who presents with
respiratory distress and hypotension. He is
emergently intubated in the ER and fluid
resuscitated with 3L of NS.
– LL has NKDA, weighs 91kg and his admit SCr=4.5
mg/dl
• His SCr at a clinic visit one month prior = 4.1 mg/dl
– Cefepime, ciprofloxacin and vancomycin are
written for – What doses should be given?
 Renal Function – Chronic Kidney Disease
• Recommendations from “A clinical update from Kidney Disease:
Improving Global Outcomes (KDIGO)”
• Delayed attainment of steady state due to reduced clearance
and prolonged half-life
– Loading dose: Recommend since goal is to rapidly achieve the
desired steady state concentration
• Especially if antibiotic has a long half-life
– Maintenance dose:
• Time dependent antibiotics: decrease the dose, but maintain the same
dosing regimen
• Concentration dependent antibiotics: give the same dose, but prolong the
dosing interval
– Therapeutic drug monitoring:
• Take into account there may be differences in unbound drug concentration
• What dose to give?

Kidney International 2011;80:1122-37

 Patient Case Continued
• The next day LL’s SCr=5.1 mg/dl and he is
anuric and on norepinephrine. The renal
consult team recommends starting renal
replacement therapy and either CRRT or
SLEDD is started.
• How do you adjust the antibiotic doses?
 Renal Function - RRT
• BIG issue with these modalities -> Lack of data
• CRRT
– Method 1: Dose as if the CrCl ~ 20-50 ml/min
– Method 2: Divide hourly ultrafiltrate rate by 60 to get estimated CrCl
• 3000 ml/hour divided by 60 = est CrCl of 50 ml/min)
– Method 3: Use general table or literature values for specific medications
• Trotman RL. CID 2005;41:1159-66
• Pea F. Clin Pharmacokinet 2007;46:997-1038
• Heintz BR. Pharmacotherapy 2009;29:562-77
– Method 4: Use an estimation formula (Curr Opin Crit Care 13:645-51)
• Total body clearance (TBC) = Clearance non-renal (CLNR) + Clearance CRRT (CLcrrt )
• SLEDD
– Method 1: (Clin Inf Dis 2009;433-7)
• If SLEDD lasts for 6-12 hours/day: dose for CRRT, namely an estimated CrCl ~10-50 ml/min
• Antibiotics dosed every 24 hours: give after SLEDD daily
• Antibiotics dosed every 12 hours: give after SLEDD and 12 hours later
• Check serum levels immediately after SLEDD to determine need for supplemental dose
– Method 2: (Crit Care Med 2011;39:560-70)
• For blood flow rate 200 ml/min and dialysate flow rate 100 ml/min, dose antibiotics for estimated CrCl
60 ml/min while on SLEDD and 10 ml/min while off SLEDD
 Revised Patient Case
• LL is a 26yo patient with a history of a MVC 7
days ago who develops respiratory distress
and hypotension on the floor. He is
emergently intubated, transferred to the ICU
and fluid resuscitated with 3L of NS.
– LL has NKDA, weighs 91kg and his current SCr=0.4
mg/dl
• His SCr on admission= 0.7 mg/dl
– Cefepime, ciprofloxacin and vancomycin are
written for – What doses should be given?
Renal Function – Augmented Renal Clearance
• Definition: CrCl value > 10% above the upper limit of normal
• At risk for subtherapeutic dosing, treatment failure and
development of resistant organisms
• Patients at risk: younger patients (~<55 years), post trauma
(especially head injuries), post-op, sepsis, burns and
hematologic malignancies
• Not a lot of data
• Recommendations:
– Use timed CrCl collections to determine renal function
– May need to use continuous infusions for beta-lactams and
vancomycin
– Use therapeutic drug monitoring when available
• What dose to give?

Clin Pharmacokinet 2010;49:1-16

 Revised Patient Case
• LL is a 45yo patient with a history of a renal
transplant in 2007 who presents with respiratory
distress and hypotension. He is emergently
intubated in the ER and fluid resuscitated with 3L
of NS.
– LL has NKDA, weighs 91kg, his admit SCr=1.2 mg/dl
and his AST=1245 U/l (nl 0-50), ALT=2312 U/l (nl 12-
78) and his tbili=1.5 mg/dl (nl 0-1.4)
– Cefepime, ciprofloxacin and vancomycin are written
for – What doses should be given?
 Hepatic Dysfunction
• Not a lot of data, especially with acute dysfunction
– No simple endogenous marker to predict function clinically used
– No available dosing adjustment tables
• Manufacturers, mostly for newer agents, have included dosing
recommendations based on Child-Pugh scores
– The FDA and European Medicines Agency (EMEA) recommend that a
kinetic study be conducted in agents that are likely to be
used/affected by hepatic dysfunction – use Child-Pugh score
• Phase 1 reactions are affected more than phase 2 reactions in
mild-to-moderate liver dysfunction
– Phase 2 reactions ARE affected by severe hepatic dysfunction
• Recommended dosing adjustments
– Depends on extraction ratio and protein binding
• What dose to give?

Eur J Clin Pharmacol 2008;64:1147-61

 Revised Patient Case
• LL is a 45yo patient with a history of a renal
transplant in 2007 who presents with
respiratory distress and hypotension. He is
emergently intubated in the ER and fluid
resuscitated with 6L of NS.
– LL has NKDA, weighs 191kg and his admit SCr=1.4
mg/dl
• His SCr at a clinic visit one month prior = 1.3 mg/dl
– Cefepime, ciprofloxacin and vancomycin are
written for – What doses should be given?
 Obesity
• Pharmacokinetic changes in obesity in general
– Absorption
• Little data exists on differences -> maybe delayed gastric emptying
– Distribution
• Lipophilic medications should be dosed on total body weight due to higher distribution volumes
• Hydrophilic medications should be dosed on ideal body weight or adjusted body weight due to
lower volumes of distribution
– Metabolism
• CYP3A4 has lower drug clearance; CYP2E1 and most phase 2 enzyme systems have higher
clearance; CYP1A2, CYP2C9, CYP2C19 and CYP2D6 trend towards higher clearance
– Excretion
• Obesity results in an increase in baseline renal clearance, but has a higher incidence of renal
dysfunction from hypertension or diabetes
• Estimate CrCl:
– Am J Health-Syst Pharm 2009;66:642-8: Cockcroft-Gault equation with fat-free weight (using
bioelectrical impedence) or lean body weight provided unbiased estimates
– Pharmacotherapy 2012;32:604-12: Obese patients (BMI 25 to >40 kg/m2), using the Cockcroft-Gault
equation with an adjusted body weight using a factor of 0.4 was the most accurate
– What dose to give?

Curr Opin Infect Dis 2012;25:634-49

Clin Pharmacokinetic 2012;51:277-304
 Conclusions
• Need to make antibiotic dosing recommendations
fast without a lot of data
• Give high normal to higher than recommended
loading doses
• In patients without organ dysfunction, give the
highest recommended dose
• In patients with organ dysfunction:
– Acute kidney dysfunction without history -> give normal
dose for 24-48 hours and monitor closely
– Acute hepatic dysfunction without history -> give normal
dose and monitor closely
 Questions?

Acknowledgements
Matt Willenborg, PharmD
Melissa Heim, PharmD
Andrew North, Pharm D
 Renal Function - CRRT
• Big issue for pharmacists with these modalities -> Lack of data
• Method 1: Dose as if the CrCl ~ 20-50 ml/min
– Concern with medications highly cleared by CRRT (i.e. fluconazole & meropenem)
• Method 2: Divide hourly ultrafiltrate rate by 60 to get estimated CrCl (i.e. 3000
ml/hour divided by 60 = est CrCl of 50 ml/min)
• Method 3: Use general table or literature values for specific medications
– Trotman RL. CID 2005;41:1159-66
– Pea F. Clin Pharmacokinet 2007;46:997-1038
– Heintz BR. Pharmacotherapy 2009;29:562-77
• Method 4: Use an estimation formula (adapted from Curr Opin Crit Care 13:645-51)
– Total body clearance (TBC) = Clearance non-renal (CLNR) + Clearance CRRT (CLcrrt )
– CLcrrt = Sieving coefficient (S) x ultrafiltrate rate + dialysis flowrate
• S = concentration drug in ultrafiltrate / concentration drug in blood
– May be estimated by fraction of drug unbound
– CLNR = Vd x elimination rate constant in HD patients (KHD)
– Fraction removed by CRRT (frcrrt) = CLcrrt / TBC
• If < 0.25: no need to supplement dose; If > 0.25: supplemental dose necessary
– Maintenance dose multiplication factor= 1/1- frcrrt
– CRRT dose = MDMF x anuric dose
• If concentration dependent drug: Increase total dose, keep same interval
• If time dependent drug: Keep same dose, change interval

Kidney International 2011;80:1122-37

 Renal Function - CRRT
• Example: Acyclovir in a 70kg person undergoing
CVVH with an UFR = 2450ml/hr
– CLcrrt = S x UFR = 0.85 x 2450ml/hr = 34.7ml/min
• Protein binding = 15%
– CLNR = Vd x KHD = 56L x 0.04hr-1 = 2.24L/hr = 37.3ml/min
• Vd = 0.8 L/kg; t1/2 = 19.5 hrs; KHD = 0.04 hr-1
– TBC = CLNR + CLcrrt= 34.7ml/min + 37.3ml/min = 72ml/min
– frcrrt = CLcrrt / TBC = 34.7ml/min / 72ml/min = 0.48
– MDMF = 1/1- frcrrt = 1/(1-0.48) = 1.92
– CRRT dose = MDMF x anuric dose = 1.92 x 5mg/kg/day
= 9.6 mg/kg/day
– Will change interval so would give: 5mg/kg IV Q12H
 Renal Function - SLEDD
• Sustained low efficient daily dialysis
• Hybrid form of dialysis that has combined advantages of intermittent HD
and CRRT
– Uses intermittent HD equipment with reduced blood and dialysate flow rate
• Usual duration is 8-12 hours/day to continuous
• Medication removal is through diffusion
• Lack of data on drug removal with this form of dialysis
• Recommendations from CID 2009:
– If SLEDD lasts for 6-12 hours/day: for renally cleared antibiotics, dose for CRRT, namely an estimated
CrCl ~10-50 ml/min
– Antibiotics dosed every 24 hours: give after SLEDD daily
– Antibiotics dosed every 12 hours: give after SLEDD and 12 hours later
– Check serum levels immediately after SLEDD to determine need for supplemental dose
• Recommendations from CCM 2011 (Nebraska Medical Center)
– For blood flow rate 200 ml/min and dialysate flow rate 100 ml/min, dose antibiotics for estimated CrCl
60 ml/min while on SLEDD and 10 ml/min while off SLEDD
» Individualize dosing based on residual renal function and whether the patient is receiving
intermittent HD

Crit Care Med 2011;39:560-70

Clin Inf Dis 2009;433-7
 Hepatic Dysfunction
• Recommended dosage adjustments
– High extraction ratio
• Oral bioavailability can be drastically increased
• Clearance may be reduced if decreased hepatic blood flow
– Low extraction ratio and high protein binding (> 90%)
• Clearance may be reduced depending on enzyme system
involved and degree of hepatic dysfunction
• Follow unbound concentrations if available
– May have high concentrations even if total concentrations are
within normal limits
– Low extraction ratio and low protein binding (< 90%)
• Clearance may be reduced depending on enzyme system
involved and degree of hepatic dysfunction
• Usually only need to follow total concentrations

Eur J Clin Pharmacol 2008;64:1147-61