Skip to main content

Table 3 Estimated effects of implementation of different empiric sepsis treatments on effective therapy rate and consumption of carbapenems in a population suspected of Gram-negative bacteremia

From: Using local clinical and microbiological data to develop an institution specific carbapenem-sparing strategy in sepsis: a nested case-control study

Treatment strategy Sensitivity of the criterion for presence of combined resistance* Proportion of patients with Gram-negative BSI adequately treated Proportion of patients with Gram-negative BSI treated with carbapenem Estimated NNTC** with carbapenem according to frequency of Gram-negative bacteremia in suspected sepsis
A priori probability of Gram-negative bacteremia in suspected sepsis a
 
     5% 10% 20% 30% 40%
1. Cefuroxime/gentamicin in all patients with sepsis 0 .912 0
2. Carbapenem in all patients with sepsis 1.000 .998 1.000 233 116 58 39 29
3. Only a carbapenem in patients with antibiotic pre-treatment on day of culture. .690 .971 .296 100 50 25 17 13
3. Only a carbapenem in patients with antibiotic treatment < 2 months .943 .993 .529 130 65 33 22 16
4. Only a carbapenem in patients with a DRPb cultured < 6 months .465 .952 .111 55 28 14 9 7
5. Only a carbapenem in patients with a DRP cultured previously (no time restriction) .592 .963 .195 76 38 19 13 10
7. Only a carbapenem in patients with a DRP previously and antibiotic treatment < 2 months .549 .961 .101 42 21 11 7 5
8. Current Practice .225 .931 .056 57 29 14 10 7
  1. Legend A Frequency of Gram-negative bacteremia as percentage of the total No. of patients with suspected sepsis in whom empiric therapy is started. B Drug resistant pathogen(s) (DRP) isolated from any body site: Vancomycin resistant enterococci, multi resistant Staphylococcus aureus, Enterobacteriaceae with in vitro resistance to aminoglycosides, second and/or third generation cephalosporin’s (including ESBL positive Enterobacteriaceae) and/or quinolones, Pseudomonas aeruginosa with resistance to third generation cephalosporins, aminoglycosides or quinolones.* The sensitivity was derived from the study data (cases 2013–2016) ** NNTC = Number needed to treat with carbapenem instead of cefuroxime/gentamicin to avoid mismatch of empiric therapy for Gram-negative bacteremia in one patient. For the calculation of the NNTC the formula in Additional file 1: Supplement A was applied
  2. Example, strategy 5: Standard empiric treatment is cefuroxime/gentamicin, carbapenems are reserved for patients with a history of drug resistant pathogen (DRP). This results in prescription of a carbapenem in 19.5% of patients with Gram-negative bacteremia. With this strategy, empiric treatment of patients with cefuroxime/gentamicin resistant bacteremia is adequate in 59.2% and the overall treatment adequacy rate in Gram-negative bacteremia is 96.3%. In the scenario of a pre-test probability of Gram-negative bacteremia of 10%, 38 patients would be treated with a carbapenem to avoid mismatch of empiric therapy for Gram-negative bacteremia in 1 patient
\