Gentamicin · Tobramycin · Amikacin · Reviewed by a PharmD, BCCCP · Updated Oct 2026
The 2023 CLSI breakpoints assume 7 mg/kg gentamicin or tobramycin and 15 mg/kg amikacin, once daily, with normal renal function, and they were set on a bacterial stasis endpoint. A smaller dose or a q8h schedule may miss the target even when the isolate reads “S.” Outside the urinary tract, use an aminoglycoside as a second agent alongside another active drug (usually a beta-lactam), not as monotherapy (CLSI M100-Ed33). All doses are adult and use the dosing weight below.
Fast Lookup (adult, CrCl ≥60 mL/min)
Drug
Extended-interval (EID)
Traditional
Traditional targets (mg/L)
Label ceiling
Gentamicin
5-7 mg/kg q24h. Hartford: 7 mg/kg.
1-1.7 mg/kg q8h (label: 3 mg/kg/day, up to 5 mg/kg/day if life-threatening)
Peak 4-10 by indication; trough <1-2
Avoid prolonged peaks >12 and troughs >2
Tobramycin
5-7 mg/kg q24h. Hartford: 7 mg/kg.
1-1.7 mg/kg q8h (label: 3 mg/kg/day, up to 5 mg/kg/day)
7.5 mg/kg q12h or 5 mg/kg q8h (label 15 mg/kg/day, max 1.5 g/day)
Peak 15-30 by indication; trough <4-8
Avoid peaks >35 and troughs >10
Synergy (gent)
Gram-positive endocarditis: 3 mg/kg/24 h, once daily (viridans strep) or in 2-3 divided doses (enterococcus, staph PVE) per AHA 2015. AHA 2026 drops most of these uses. See endocarditis▼.
Divided dosing: peak 3-4, trough <1
Gent is not used for staph native valve IE
EID: IDSA/ATS HAP/VAP 2016 ranges (as tabulated in Bland, Pharmacotherapy 2018). Traditional doses and targets by indication follow a representative academic protocol (Stanford Health Care 2021); targets vary by institution. Label values from US gentamicin, tobramycin, and amikacin prescribing information. Septic shock amikacin: French guideline dose cited in Allou, Ann Intensive Care 2016. Labeled duration is 7-10 days; amikacin safety beyond 14 days is not established. Label peak ceilings apply to traditional divided dosing; EID and septic shock doses intentionally exceed them, and the amikacin label caps traditional dosing at 15 mg/kg/day.
Dosing Weight
Patient
Weight to dose on
Actual weight below IBW
Actual (total) body weight
Actual weight up to ~120% of IBW
Actual body weight for EID (Hartford). Many protocols use IBW for traditional dosing.
Actual weight >120% of IBW (1.2 × IBW)
Adjusted weight = IBW + 0.4 × (actual − IBW). The tobramycin label states the same rule: lean weight plus 40% of the excess.
Endocarditis synergy
AHA 2015 doses gentamicin on ideal body weight.
💡
The obesity trigger differs by protocol: >20%, >30% above IBW, or BMI ≥30 are all in use. The 0.4 factor is consistent. Use your institution’s threshold and the same weight for Cockcroft-Gault. IBW (Devine): men 50 kg, women 45.5 kg, plus 2.3 kg per inch over 5 feet.
7 mg/kg gentamicin or tobramycin on dosing weight, infused over 30-60 min. The nomogram was built for this dose only; it misreads clearance at 5 mg/kg or any other dose.
2
Initial interval by CrCl: ≥60 mL/min q24h · 40-59 q36h · 20-39 q48h · <20 do not use.
3
Draw one random level 6-14 h after the START of the first infusion (some protocols use 8-12 h). Record the actual draw time.
4
Plot level against hours since infusion start. The zone sets the interval: q24h, q36h, or q48h.
5
Level falls on a line: choose the longer interval.
6
Level above the q48h zone: stop scheduled dosing. Switch to level-guided (traditional) dosing; redose only when levels show clearance.
7
Repeat a random level about weekly, and any time renal function changes.
8
Amikacin 15 mg/kg: many protocols halve the amikacin level before plotting. That is a local extension; the 1995 cohort was 94% gentamicin.
⚠️
Do not use the nomogram in: CrCl <20 mL/min (many protocols use <30) or unstable renal function, any dialysis or CRRT, pregnancy, burns >20% TBSA, ascites, anasarca or other large fluid shifts, gram-positive synergy or endocarditis, cystic fibrosis (CF-specific dosing), and pediatrics. In these patients Vd or clearance is too far from the validation population for one level to place the patient correctly.
💡
Why EID is the default: in 2,184 patients the Hartford program had 1.2% nephrotoxicity against a 3-5% historical rate (Nicolau 1995). Across 21 randomized trials, once-daily dosing was as effective as multiple daily dosing with lower nephrotoxicity (RR 0.74, 95% CI 0.54-1.00) and no more ototoxicity (Barza, BMJ 1996).
Two-Level Individualized Kinetics (Sawchuk-Zaske) ↑ back to top
Use when the nomogram does not apply, when the MIC is known, or when you want a measured Cmax. One-compartment, first-order model (Sawchuk and Zaske, J Pharmacokinet Biopharm 1976). Bayesian software does the same job from 1-2 levels without steady state and is preferred where available.
Step
Do this
1
Draw C1 about 1-2 h after the end of infusion (post-distribution) and C2 about 8-12 h after the start of the same dose.
True Cmax at end of infusion: Cmax = C1 × e(ke × t), where t = hours from end of infusion to C1.
4
Volume (no residual pre-dose level): Vd = Dose × (1 − e−ketinf) ÷ (tinf × ke × Cmax).
5
New dose for a target Cmax: Dose = Cmaxtarget × Vd × ke × tinf × (1 − e−keτ) ÷ (1 − e−ketinf), using τ from step 6. The (1 − e−keτ) term corrects for accumulation; it is close to 1 only when the trough is near zero. Target Cmax/MIC 8-10.
6
Interval: τ = tinf + ln(Cmaxtarget ÷ Cmintarget) ÷ ke (Cmin target typically <1 mg/L for gent/tobra). Round to a practical interval.
💡
Where 8-10 comes from: clinical response was about 92% at a maximal Cmax/MIC ≥10, 88% at 8 to <10, and about 70% at 4 to <6 (Moore, J Infect Dis 1987). In gram-negative nosocomial pneumonia, Cmax/MIC ≥10 within 48 h gave a 90% probability of fever and WBC resolution by day 7 (Kashuba, AAC 1999).
⚠️
Peak timing matters. A sample in the distribution phase overstates Cmax; one series found a 30-minute level about twice the 2-hour level (Bland 2018). AUC-based targets (AUC/MIC about 30-50 for low-burden or combination therapy, 80-100 for critically ill or monotherapy) are proposed but not standardized.
Traditional Dosing Targets by Indication
Indication
Gent / tobra peak
Amikacin peak
Trough
Life-threatening (pneumonia, septic shock)
8-10
25-30
Gent/tobra <1-2; amikacin <4-8 (aim for the low end when possible)
Serious systemic infection
6-8
20-25
UTI
4-6
15-20
Gram-positive synergy (divided)
3-4
Not used
<1
mg/L. Peak 30 min after end of a 30-min infusion; trough just before the next dose; draw around the 3rd-4th dose. Indication targets: Stanford Health Care protocol (2021); synergy targets: AHA 2015. Numbers vary between institutions by about one band.
Amikacin: When 15 mg/kg Is Not Enough
Setting
Dose
Target
Septic shock, first dose
25-30 mg/kg (French guidelines)
Cmax 60-80 mg/L (Cmax/MIC ~8-10). In one ICU cohort only 42% landed in range at 30 mg/kg (Allou 2016).
NTM lung disease, daily
10-15 mg/kg IV daily
Peak 35-45; trough <5
NTM lung disease, intermittent
15-25 mg/kg IV three times weekly
Peak 65-80; trough <5
Refractory MAC
Amikacin liposome inhalation 590 mg daily
Not a systemic regimen
NTM: ATS/ERS/ESCMID/IDSA 2020 guideline. About one third of patients developed permanent ototoxicity over 15 weeks on these regimens, tied to age and cumulative dose; the guideline favors lower doses and intermittent dosing.
Gram-Positive Endocarditis Synergy (AHA 2015 doses, AHA 2026 update) ↑ back to top
⚠️
AHA replaced its 2015 statement in September 2026 (DeSimone, Circulation 2026) and narrowed gentamicin use. As summarized at publication: gentamicin is not advised in staph PVE (toxicity outweighs benefit); VGS with intermediate penicillin susceptibility gets ceftriaxone alone; gentamicin for the first 2 weeks with ceftriaxone is kept for fully penicillin-resistant VGS; combination therapy is supported for E. faecalis, with ampicillin + ceftriaxone × 6 weeks favored over ampicillin + gentamicin. One summary also reports that the 2-week gentamicin short course for highly susceptible VGS was dropped. ESC 2023 still lists gentamicin regimens (2 weeks with ampicillin for E. faecalis; 2 weeks in staph PVE). The table gives the AHA 2015 doses for when gentamicin is still chosen; confirm against the 2026 full text and your ID team.
Organism / valve
Gentamicin
Notes
Viridans strep, native valve, PCN MIC ≤0.12
3 mg/kg/24 h once daily × 2 weeks with penicillin or ceftriaxone (2-week regimen)
Uncomplicated, low-risk patients only. Not with CrCl <20, extracardiac infection, or Abiotrophia, Granulicatella, Gemella.
Viridans strep, native valve, MIC >0.12 to <0.5
3 mg/kg/24 h once daily for the first 2 weeks of a 4-week beta-lactam course
Once daily is preferred; if given q8h, target peak 3-4, trough <1. AHA 2026: ceftriaxone alone when ceftriaxone MIC ≤0.5.
Viridans strep, prosthetic valve
3 mg/kg/24 h × 2 weeks (optional if PCN MIC ≤0.12); reasonable to extend to 6 weeks if MIC >0.12
No cure benefit shown over beta-lactam alone. Do not give if CrCl <30.
Enterococcus (PCN- and gent-susceptible)
3 mg/kg IBW/24 h in 2-3 equal doses (about 1 mg/kg q8h) with ampicillin 2 g q4h; native valve 4 weeks if symptoms <3 months, 6 weeks if longer; prosthetic valve 6 weeks
Peak 3-4, trough <1. Ampicillin 2 g q4h + ceftriaxone 2 g q12h × 6 weeks is recommended if CrCl is <50 at baseline or falls on gentamicin, and is reasonable at any renal function (AHA 2015). AHA 2026 favors it as the default.
Staph, prosthetic valve
AHA 2015 and ESC 2023: 3 mg/kg/24 h × 2 weeks (AHA: in 2-3 doses), with nafcillin/oxacillin or vancomycin plus rifampin. AHA 2026: do not add gentamicin.
Beta-lactam (or vancomycin) continues at least 6 weeks. AHA 2026 left rifampin to the endocarditis team.
Staph, native valve (MSSA or MRSA)
Do not add gentamicin (Class III)
Includes right-sided IE in people who inject drugs.
💡
The field is moving away from aminoglycosides here. The 2023 WikiGuidelines consensus (JAMA Netw Open) states that most authors suggest against routine adjunctive aminoglycosides because of harm without proven benefit. AHA 2026 now advises against gentamicin in staph PVE (AHA 2015 allowed substituting a susceptible aminoglycoside or omitting it). Ampicillin plus ceftriaxone covers high-level aminoglycoside-resistant E. faecalis.
Renal Dysfunction, Hemodialysis, CRRT
Setting
Gentamicin / tobramycin
Amikacin
CrCl 20-59 (EID)
Keep the mg/kg dose and extend the interval (q36h-q48h), then confirm with a level. CrCl <20: traditional dosing guided by levels.
Intermittent HD
Load 2 mg/kg, then 1.5 mg/kg after each HD (label range 1-1.7 mg/kg post-HD)
Pre-HD level by indication (Heintz 2009: redose when <10)
CRRT
Load 2-3 mg/kg, then 1.5-2.5 mg/kg q24-48h by levels (systemic gram-negative; Heintz lists 1 mg/kg q24-36h for mild UTI or synergy)
Load 10 mg/kg, then 7.5 mg/kg q24-48h by levels
Stanford Health Care protocol (2021) and Heintz, Pharmacotherapy 2009; HD label range from the gentamicin PI. Always give the dose after the dialysis session. In AKI, Cockcroft-Gault lags true clearance; use early levels instead of the calculated CrCl.
⚠️
These renal doses were built for traditional peak targets. Against an Enterobacterales isolate at the new MIC of 2 mg/L, a 1.5 mg/kg dose will not approach Cmax/MIC 8-10. For a serious infection in a dialysis patient, check the MIC and dose to a measured peak, ideally with ID or pharmacokinetics input.
Toxicity Monitoring
Toxicity
What to know
Monitor
Nephrotoxicity
10-25% in most studies, 29-58% in critically ill cohorts. Classically non-oliguric, often reversible, appearing after about 7-10 days. Risk: volume depletion, sepsis, baseline CKD, longer courses, older age, liver disease, and concurrent nephrotoxins (vancomycin and vasopressors were the strongest predictors in one SICU cohort).
SCr at least every 48 h; daily in ICU or unstable renal function. Rising SCr: recheck a level before the next dose.
Ototoxicity
Cochlear and vestibular; often permanent. Dose- and duration-related, but carriers of MT-RNR1 variants (m.1555A>G, m.1494C>T, m.1095T>C) can lose hearing after one dose with levels in range. m.1555A>G is present in about 0.2% of people in the US.
Ask about hearing loss, vertigo, and maternal family history of aminoglycoside deafness. Consider baseline and serial audiometry for courses >2 weeks.
Neuromuscular blockade
Boxed warning on the amikacin label and a labeled warning for all three; can cause respiratory paralysis, especially with neuromuscular disease or paralytics.
Infuse over at least 30 min.
💡
CPIC (2021): known MT-RNR1 carriers should avoid aminoglycosides unless the infection risk outweighs the risk of permanent hearing loss. If no alternative exists, use the shortest course, with TDM and frequent hearing checks. The gentamicin label also says to avoid potent diuretics (furosemide, ethacrynic acid) and other neuro- or nephrotoxins, and that aminoglycosides can cause fetal harm.
mg/L. Based on 7 mg/kg gent/tobra and 15 mg/kg amikacin q24h and a net-stasis target. FDA completed its review and recognized these breakpoints in October 2025.
💡
What it means at the bedside: (1) Aminoglycoside monotherapy is for urinary infections; elsewhere it is an add-on. (2) Do not count on gentamicin for Pseudomonas; tobramycin is the aminoglycoside with a systemic breakpoint. (3) Amikacin lost the most: in SENTRY isolates, susceptibility among ESBL producers fell from 96.9% to 79.7%, and among CRE from 75.2% to 59.0% (Sader, OFID 2023). (4) Labs adopt new breakpoints on different timelines. Ask your lab which breakpoints it reports before you read an old antibiogram.
Using the Hartford nomogram outside its population. In dialysis, CRRT, CrCl <20, burns, ascites, or pregnancy, a single random level puts the patient in the wrong zone. Use traditional, level-guided dosing.
☠️
Plotting the wrong clock time or the wrong dose. The x-axis is hours from the start of infusion. A 5 mg/kg dose plotted on the 7 mg/kg nomogram, or an amikacin level plotted unadjusted, assigns the wrong interval.
☠️
Dosing an obese patient on total body weight (overdose), or rounding a CrCl from an AKI creatinine (non-steady state) into a q24h interval. Recheck a level early when SCr is moving.
☠️
Aminoglycoside alone for bacteremia or pneumonia, or gentamicin for Pseudomonas. The 2023 breakpoints do not support either.
☠️
Stacking nephrotoxins and running long courses. Aminoglycoside plus vancomycin plus a loop diuretic or contrast, past 7 days, with SCr checked twice weekly, is the classic path to AKI and permanent hearing loss. Reassess the need for the aminoglycoside daily.