First doses of aminoglycoside and -lactam antibiotics, when used in combination, are usually given simultaneously; however, nonsimultaneous administration may be more efficacious. We used a dynamic in vitro model, which simulates in vivo serum kinetics, to assess the effect of spacing the first doses of gentamicin and ceftazidime used against Pseudomonas aeruginosa ATCC 27853 and two clinical isolates of P. aeruginosa, PA1 and PA2. The following dose regimens against P. aeruginosa ATCC 27853 were compared: (i) gentamicin given alone, (ii) ceftazidime given alone, (iii) gentamicin and ceftazidime given simultaneously, (iv) gentamicin followed by ceftazidime at 15 or 50 min or at 2, 4, or 8 h, and (v) ceftazidime which was followed by gentamicin at 4 h. The effects of regimen iii and the 4-h interval in regimen iv against PA1 and PA2 were also compared. Initial peak concentrations used were 8 mg/liter for gentamicin and 80 mg/liter for ceftazidime, with drug half-lives of 2.5 and 1.8 h, respectively. Compared with simultaneous administration, nonsimultaneous administration (regimens iv and v) produced greater overall bacterial killing and was associated with a delay in bacterial regrowth (P < 0.005) of up to 6.6 to 8.3 h, regardless of the order in which the drugs were given. The optimal interval between gentamicin and ceftazidime doses, which maximized initial bactericidal effect and the time before regrowth, appeared to be 2 to 4 h.When intravenous antibiotics are given in combination for the treatment of serious infections, the first doses of the two drugs are usually given simultaneously. This is a convenient way of administration, and there has been no reason to suspect that simultaneous administration results in less than optimal bactericidal effect. We have used a dynamic in vitro model to explore variations in dosing regimens of individual antibiotics and combinations to determine whether novel regimens might have greater bactericidal activity than standard regimens (1, 2, 4). In 1986, König et al. used a similar model to show that gentamicin and ampicillin were more effective against Escherichia coli when given 4 h apart than when administered simultaneously (10). The same group later showed the same phenomenon with gentamicin and ticarcillin against Pseudomonas aeruginosa (8). In both cases, overall bacterial killing was greater and there was a delay in bacterial regrowth with nonsimultaneous administration, regardless of the order in which the drugs were given. Intervals other than 4 h were not explored, and, to our knowledge, other antibiotic combinations have not been tested to determine whether this is a generalized phenomenon with aminoglycosides and -lactam antibiotics.In this study, we have investigated the combination of gentamicin and ceftazidime against three strains of P. aeruginosa, comparing spaced administration with simultaneous administration and varying the time interval between doses to determine if there was an optimal regimen. Use of the dynamic in vitro model enabled simulation of ...