Clinical and Economic Evaluation of Oral Ciprofloxacin After an Abbreviated Course of Intravenous Antibiotics JOSEPHA. PALADINO, Pharm.D., HOWARD E. SPERRY, M.D., JULIE M. BACKES, Pharm.D., JEFFREYA. GELBER, M,D., DEBORAH J. SERRIANNE, R.N., THOMAS J. CUMBO, M.D., JEROME J. SCHENTAG, Pharm.D., Buffalo, New York
PURPOSF~ Oral ciprofloxacin has the requisite pharmaeoklnetic and antibacterial properties to rival the potency of intravenous antibiotics. This s t u d y was designed to determine whether oral ciprofloxacin could abbreviate the course of intravenons antibiotics in the treatment of serious infections. PATIENTS AND METHODS: HoopitaliT~l a d u l t patients were eligible for enrollment if they had a
serious infection that was expected to require 8 or more days of intravenous antibiotic treatment. After conventional intravenous antibiotics were administered for 3 days, informed consent was obtained and patients were randomly assigned to either continue parenteral antibiotics (n = 53) or switch to oral ciprofloxacin 750 mg taken twice daily (n = 52). Ninety-nine of the 105 patients were evaluable for the assessment of efficacy. Clinical and bacteriologic efficacy, adverse events, and coots of the two treatments w e r e compared. RESULTS: The two treatment groups were comparable for demographic characteristics, types of infectious, bacteria isolated, initial intravenous antibiotic regimens, and duration of antibiotic treatment. The moot common infections were of the skin and skin structure; bacteremia and infections of the lower respiratory tract, urinary tract, and bone and joint were also represented. The moot commonly isolated pathogens were Staphylococcus aureus, Pseudomon~ aeruginosa, and Escherichia coll. The moot frequently prescribed intravenous antibiotics before randomiT~tion included nmlnoglycooides, cephalooporins, vancomycin, and nsfciillu; 52 evaluable patients were treated with combination therapy while 47 received monotherapy. The clinical and bacteriologic outcomes and adFrom the Clinical PharmacokineticsLaboratory (JAP, JMB, DJS, JJS) and the Department of Medicine(HES,JAG,TJC). Millard FillmoreHospitals, and the State University of New York at Buffalo, Buffalo, New York. This work was supported in part by a researchgrant from Miles, Inc. Requests for reprints should be addressed to Joseph A. Paladino, Pharm.D., Millard Fillmore Suburban Hospital, 1540 Maple Road, Williamsville,New York 14221. Manuscript submitted February 14. 1991, and accepted in revised form July 11, 1991.
462
verse reaction frequency with oral ciprofloxacin were comparable to those of the continued intravenous antibiotic regimens. Ciprofloxacin was associated with an average coot savings of $293 per patient. CONCLUSION:When used after 3 days of intravenous antibiotics, oral ciprofloxacin was as safe and effective as full courses of intravenous antibiotics and provided substantial coot savings.
tandard treatment for serious infections in-
S cludes hospitalization and the administration of intravenous antibiotics for 7 to 10 days or longer. Typically, the length of treatment for the intravenous route is empirically chosen or based on published standards. As proposed by Quintiliani et al [1], one may question whether the full course of intravenous antibiotics is necessary and whether it can be abbreviated by switching to potent oral agents after a shorter duration of intensive treatment with parenteral agents. The fluoroquinolone antibiotics are notable for potent oral activity against a broad spectrum of gram-positive and gram-negative aerobic bacteria [2]. Ciprofloxacin is one of the most potent agents of this class, having minimal inhibitory concentrations of less than 0.1 pg/mL against most Enterobacteriaceae [3,4]. This antibiotic also has activity against pathogens such as Staphylococcus aureus and Pseudomonas aeruginosa [3,4]. Pharmacokinetic studies of oral ciprofloxacin document reliable bioavailability [5] with resultant serum and tissue concentrations in excess of those required to eradicate susceptible bacterial pathogens [6-8]. Compared with intravenous antibiotics, oral medications are more comfortable and convenient for the patient to receive, require minimal expertise or time and few ancillary materials to prepare and administer, and cost less to procure [9-11]. Accordingly, a hypothesis was developed that orally administered ciprofloxacin could be utilized in the treatment of serious infections in the hospitalized patient. Because seriously ill patients are frequently incapable of ingesting or adequately absorbing oral medications [12], the protocol required an ini-
November 1991 The American Journal of Medicine Volume 91
ORAL ClPROFLOXAClN VS INTRAVENOUSANTIBIOTICS / PALADINO ET AL
tial course of 3 days of intravenous antibiotic treatment. Based on the nature and severity of the infection, patients were expected to be treated with intravenous antibiotics for a minimum duration of 8 days. Therapy with oral ciprofloxacin commenced on Day 3 for those patients who were randomly assigned to the oral treatment group. Thus, the efficacy, safety, and costs of a full course of intravenous antibiotics were compared with oral ciprofloxacin employed as continuation therapy.
given oral antibiotics, but if the minimum of 8 days of parenteral antibiotics was not achieved, then the patient was not evaluable. Patients in this control group received optimal standard care, which included outpatient parenteral antibiotic therapy when indicated. Modifications to the intravenous regimens were allowed for dosage adjustment or regimen streamlining. If antibiotic therapy was changed because of an unsatisfactory response, treatment was considered a failure.
PATIENTS AND METHODS Patients
Laboratory Studies
Hospitalized adults were eligiblefor enrollment if they had a serious bacterial infection expected to require at least 8 days of intravenous antibiotic treatment and were able to ingest oral medications by the time of the randomization procedure. Isolated pathogens had to be susceptible to the intravenous antibiotics and to ciprofloxacin.Pregnant or lactating w o m e n and patients having any of the following conditions were excluded: history of allergic reactions to quinolones, serum creatinine level greater than 3.0 mg/dL (265 #tool/L), chronic indwelling urinary catheter, neutropenia (less than 1,000 cells/ram3),meningitis, endocarditis,cysticfibrosis, anaerobic bacteremia, or pelvic infections. The study protocol was approved by the Investigational Review Board of Millard Fillmore Hospitals, and written informed consent was obtained from each patient prior to enrollment.
Treatment Protocol All patients were initiallytreated with intravenous antibioticsas selected by their attending physicians. Any antibiotic regimen that was clinically ineffective or bacteriologicallyunsuitable was adjusted to provide appropriate therapy. After receiving 72 hours of treatment with effectiveintravenous antibiotics,patients were randomly assigned to either continue receiving intravenous antibiotics or to switch to oral ciprofloxacin 750 m g twice daily. Randomization was not infection-specificand was performed using a singlecomputer-generated table. Oral metronidazole was added to the regimen of patients receiving ciprofloxacin where there was a need for concomitant anaerobic activity [13]. To be included in the efficacy evaluation, a patient must have received a m i n i m u m of 8 days of antibiotics; there was no restriction on the maxim u m duration of treatment. Therefore, patients randomly assigned to receive ciprofloxacinreceived 3 days of appropriate intravenous antibiotics followed by at least 5 days of oral ciprofloxacin therapy. Patients randomly assigned to continue treatment with intravenous antibiotics could also be
Cultures of the infection site,blood chemistries, hematology, and urinalysis were obtained before treatment and at randomization. These tests were repeated during the immediate posttreatment period and at follow-up 2 to 6 weeks after the last antibiotic dosage. Quantitation of ciprofloxacin in plasma was performed at steady state using a previously validated reverse-phase high-pressure liquid chromatographic assay [14]. Chromatography was performed on a Whatman Partisil 5, ODS: 3, RAC II 4.6 mm ID × 100 mm column (Whatman Incorporated, Clifton, New Jersey). Samples were detected at 440 nm with an excitation wavelength of 270 nm. The mobile phase consisted of a mixture of acetonitrile and methanol in a citric acid, ammonium perchlorate, tetrabutylammonium hydroxide buffer (final pH 2.4; flow rate = 1.3 mL/minute). Plasma proteins were precipitated by adding 300 uL of 10% perchloric acid in acetonitrile to 500 ~L of plasma to which 100 ~L of internal standard had been added (7.5 #g/mL, Abbott 56619; Abbott Laboratories, Chicago, Illinois). The plasma standard curve was linear over the range of 0.017 to 10.0 ~g/mL. Absolute recoveries of ciprofloxacin and the internal standard were determined to be 115% and 98%, respectively. The assay had a limit of quantitation and limit of detection (S/N -- 5) of 0.05 and 0.14, respectively. The average interday precision (%CV) of plasma standards for this study was 3.4% (n = 4 days over i year) with variables of 2.9, 3.3, and 6.5% for the 7.5, 1.5, and 0.15 #g/mL plasma quality controls, respectively. Long-term sample stability has been demonstrated for 2 years in plasma.
Compliance Oral administration of potent antibiotics may allow some patients to be discharged from the hospital early to complete the course of antibiotic therapy at home. The decision to treat serious infections with the unsupervised administration of potent oral antibiotics is complicated by the question of whether the patient will comply with the prescribed regimen. To address this issue, we measured compli-
November 1991 The American Journal of Medicine Volume 91
463
ORALCIPROFLOXACINVS INTRAVENOUSANTIBIOTICS/ PALADINOET AL
ance electronically using Medication Event Monitoring Systems ( M E M S : A P R E X Corporation, Palo Alto, California). M E M S monitors are normal-appearing prescription vials that contain a microprocessor in the cap to record the date and time of vial openings. Other actions to foster and monitor compliance in the study patients included specific medication counseling and detailed questioning, witnessed tablet counts, and measurement of ciprofloxacin serum concentrations. Evaluation of Therapy
Clinical efficacy and antibacterial action were assessed at randomization, during the immediate posttreatment period, and at follow-up 2 to 6 weeks after the last antibiotic dosage. Bacteriologic efficacy against the original pathogens was rated as eradication, persistence, eradication with recurrence, or indeterminate. A superinfection was recorded when a new pathogen was isolated concurrently with symptoms of infection. Clinical response was determined for each infection site at the end of therapy and rated as resolution, improvement, failure, or indeterminate. Temperature, white blood cell count with differential, and chills were monitored for each patient. In addition, the following were monitored for each infection site: (1) skin and skin structure: purulent drainage, induration, erythema, inflammation, pain, and ulceration; (2) lower respiratory tract: quantity and quality of sputum, productive cough, and chest roentgenogram; (3) urinary tract: frequency, dysuria, pyuria, and flank pain; (4) bacteremia: mental status and blood pressure; and (5) bone and joint: radiologic findings, erythrocyte sedimentation rate, purulent drainage, and localized pain, erythema, and inflammation. A n overall assessment of antibiotic treatment was made for each patient and rated as satisfactory, unsatisfactory, or indeterminate. The unsatisfactory rating included clinicalfailure as well as premature discontinuation of antibiotic secondary to an adverse event. T w o of the investigators reviewed each case to ensure consistency of evaluations. Statistical Analysis
Statistical calculations were performed by an independent group at the University of Minnesota using a Control Data Corporation Cyber 170/825 (Control Data Corporation, Bloomington, Minnesota) computer with S P S S software. S u m m a r y statistics (including means and standard deviations) were obtained for continuous and ordered variables, while frequency distributions were determined for categorical items. The two treatment groups were compared for baseline characteristics,
464
as well as antibiotic efficacy and safety. All categorical items were analyzed using chi-square with Yates' correction. Fisher's exact test was used when the sample sizewas lessthan 21. Variables that were not normally distributed underwent nonparametric analysis using the Mann-Whitney and KruskalWallis tests. Power calculations were performed to assess the potential for type II error. Unless otherwise noted, data are presented as mean ± standard deviation with the range in parentheses. All statistical analyses were conducted as two-tailed testswith p
PURPOSF~ Oral ciprofloxacin has the requisite pharmaeoklnetic and antibacterial properties to rival the potency of intravenous antibiotics. This s t u d y was designed to determine whether oral ciprofloxacin could abbreviate the course of intravenons antibiotics in the treatment of serious infections. PATIENTS AND METHODS: HoopitaliT~l a d u l t patients were eligible for enrollment if they had a
serious infection that was expected to require 8 or more days of intravenous antibiotic treatment. After conventional intravenous antibiotics were administered for 3 days, informed consent was obtained and patients were randomly assigned to either continue parenteral antibiotics (n = 53) or switch to oral ciprofloxacin 750 mg taken twice daily (n = 52). Ninety-nine of the 105 patients were evaluable for the assessment of efficacy. Clinical and bacteriologic efficacy, adverse events, and coots of the two treatments w e r e compared. RESULTS: The two treatment groups were comparable for demographic characteristics, types of infectious, bacteria isolated, initial intravenous antibiotic regimens, and duration of antibiotic treatment. The moot common infections were of the skin and skin structure; bacteremia and infections of the lower respiratory tract, urinary tract, and bone and joint were also represented. The moot commonly isolated pathogens were Staphylococcus aureus, Pseudomon~ aeruginosa, and Escherichia coll. The moot frequently prescribed intravenous antibiotics before randomiT~tion included nmlnoglycooides, cephalooporins, vancomycin, and nsfciillu; 52 evaluable patients were treated with combination therapy while 47 received monotherapy. The clinical and bacteriologic outcomes and adFrom the Clinical PharmacokineticsLaboratory (JAP, JMB, DJS, JJS) and the Department of Medicine(HES,JAG,TJC). Millard FillmoreHospitals, and the State University of New York at Buffalo, Buffalo, New York. This work was supported in part by a researchgrant from Miles, Inc. Requests for reprints should be addressed to Joseph A. Paladino, Pharm.D., Millard Fillmore Suburban Hospital, 1540 Maple Road, Williamsville,New York 14221. Manuscript submitted February 14. 1991, and accepted in revised form July 11, 1991.
462
verse reaction frequency with oral ciprofloxacin were comparable to those of the continued intravenous antibiotic regimens. Ciprofloxacin was associated with an average coot savings of $293 per patient. CONCLUSION:When used after 3 days of intravenous antibiotics, oral ciprofloxacin was as safe and effective as full courses of intravenous antibiotics and provided substantial coot savings.
tandard treatment for serious infections in-
S cludes hospitalization and the administration of intravenous antibiotics for 7 to 10 days or longer. Typically, the length of treatment for the intravenous route is empirically chosen or based on published standards. As proposed by Quintiliani et al [1], one may question whether the full course of intravenous antibiotics is necessary and whether it can be abbreviated by switching to potent oral agents after a shorter duration of intensive treatment with parenteral agents. The fluoroquinolone antibiotics are notable for potent oral activity against a broad spectrum of gram-positive and gram-negative aerobic bacteria [2]. Ciprofloxacin is one of the most potent agents of this class, having minimal inhibitory concentrations of less than 0.1 pg/mL against most Enterobacteriaceae [3,4]. This antibiotic also has activity against pathogens such as Staphylococcus aureus and Pseudomonas aeruginosa [3,4]. Pharmacokinetic studies of oral ciprofloxacin document reliable bioavailability [5] with resultant serum and tissue concentrations in excess of those required to eradicate susceptible bacterial pathogens [6-8]. Compared with intravenous antibiotics, oral medications are more comfortable and convenient for the patient to receive, require minimal expertise or time and few ancillary materials to prepare and administer, and cost less to procure [9-11]. Accordingly, a hypothesis was developed that orally administered ciprofloxacin could be utilized in the treatment of serious infections in the hospitalized patient. Because seriously ill patients are frequently incapable of ingesting or adequately absorbing oral medications [12], the protocol required an ini-
November 1991 The American Journal of Medicine Volume 91
ORAL ClPROFLOXAClN VS INTRAVENOUSANTIBIOTICS / PALADINO ET AL
tial course of 3 days of intravenous antibiotic treatment. Based on the nature and severity of the infection, patients were expected to be treated with intravenous antibiotics for a minimum duration of 8 days. Therapy with oral ciprofloxacin commenced on Day 3 for those patients who were randomly assigned to the oral treatment group. Thus, the efficacy, safety, and costs of a full course of intravenous antibiotics were compared with oral ciprofloxacin employed as continuation therapy.
given oral antibiotics, but if the minimum of 8 days of parenteral antibiotics was not achieved, then the patient was not evaluable. Patients in this control group received optimal standard care, which included outpatient parenteral antibiotic therapy when indicated. Modifications to the intravenous regimens were allowed for dosage adjustment or regimen streamlining. If antibiotic therapy was changed because of an unsatisfactory response, treatment was considered a failure.
PATIENTS AND METHODS Patients
Laboratory Studies
Hospitalized adults were eligiblefor enrollment if they had a serious bacterial infection expected to require at least 8 days of intravenous antibiotic treatment and were able to ingest oral medications by the time of the randomization procedure. Isolated pathogens had to be susceptible to the intravenous antibiotics and to ciprofloxacin.Pregnant or lactating w o m e n and patients having any of the following conditions were excluded: history of allergic reactions to quinolones, serum creatinine level greater than 3.0 mg/dL (265 #tool/L), chronic indwelling urinary catheter, neutropenia (less than 1,000 cells/ram3),meningitis, endocarditis,cysticfibrosis, anaerobic bacteremia, or pelvic infections. The study protocol was approved by the Investigational Review Board of Millard Fillmore Hospitals, and written informed consent was obtained from each patient prior to enrollment.
Treatment Protocol All patients were initiallytreated with intravenous antibioticsas selected by their attending physicians. Any antibiotic regimen that was clinically ineffective or bacteriologicallyunsuitable was adjusted to provide appropriate therapy. After receiving 72 hours of treatment with effectiveintravenous antibiotics,patients were randomly assigned to either continue receiving intravenous antibiotics or to switch to oral ciprofloxacin 750 m g twice daily. Randomization was not infection-specificand was performed using a singlecomputer-generated table. Oral metronidazole was added to the regimen of patients receiving ciprofloxacin where there was a need for concomitant anaerobic activity [13]. To be included in the efficacy evaluation, a patient must have received a m i n i m u m of 8 days of antibiotics; there was no restriction on the maxim u m duration of treatment. Therefore, patients randomly assigned to receive ciprofloxacinreceived 3 days of appropriate intravenous antibiotics followed by at least 5 days of oral ciprofloxacin therapy. Patients randomly assigned to continue treatment with intravenous antibiotics could also be
Cultures of the infection site,blood chemistries, hematology, and urinalysis were obtained before treatment and at randomization. These tests were repeated during the immediate posttreatment period and at follow-up 2 to 6 weeks after the last antibiotic dosage. Quantitation of ciprofloxacin in plasma was performed at steady state using a previously validated reverse-phase high-pressure liquid chromatographic assay [14]. Chromatography was performed on a Whatman Partisil 5, ODS: 3, RAC II 4.6 mm ID × 100 mm column (Whatman Incorporated, Clifton, New Jersey). Samples were detected at 440 nm with an excitation wavelength of 270 nm. The mobile phase consisted of a mixture of acetonitrile and methanol in a citric acid, ammonium perchlorate, tetrabutylammonium hydroxide buffer (final pH 2.4; flow rate = 1.3 mL/minute). Plasma proteins were precipitated by adding 300 uL of 10% perchloric acid in acetonitrile to 500 ~L of plasma to which 100 ~L of internal standard had been added (7.5 #g/mL, Abbott 56619; Abbott Laboratories, Chicago, Illinois). The plasma standard curve was linear over the range of 0.017 to 10.0 ~g/mL. Absolute recoveries of ciprofloxacin and the internal standard were determined to be 115% and 98%, respectively. The assay had a limit of quantitation and limit of detection (S/N -- 5) of 0.05 and 0.14, respectively. The average interday precision (%CV) of plasma standards for this study was 3.4% (n = 4 days over i year) with variables of 2.9, 3.3, and 6.5% for the 7.5, 1.5, and 0.15 #g/mL plasma quality controls, respectively. Long-term sample stability has been demonstrated for 2 years in plasma.
Compliance Oral administration of potent antibiotics may allow some patients to be discharged from the hospital early to complete the course of antibiotic therapy at home. The decision to treat serious infections with the unsupervised administration of potent oral antibiotics is complicated by the question of whether the patient will comply with the prescribed regimen. To address this issue, we measured compli-
November 1991 The American Journal of Medicine Volume 91
463
ORALCIPROFLOXACINVS INTRAVENOUSANTIBIOTICS/ PALADINOET AL
ance electronically using Medication Event Monitoring Systems ( M E M S : A P R E X Corporation, Palo Alto, California). M E M S monitors are normal-appearing prescription vials that contain a microprocessor in the cap to record the date and time of vial openings. Other actions to foster and monitor compliance in the study patients included specific medication counseling and detailed questioning, witnessed tablet counts, and measurement of ciprofloxacin serum concentrations. Evaluation of Therapy
Clinical efficacy and antibacterial action were assessed at randomization, during the immediate posttreatment period, and at follow-up 2 to 6 weeks after the last antibiotic dosage. Bacteriologic efficacy against the original pathogens was rated as eradication, persistence, eradication with recurrence, or indeterminate. A superinfection was recorded when a new pathogen was isolated concurrently with symptoms of infection. Clinical response was determined for each infection site at the end of therapy and rated as resolution, improvement, failure, or indeterminate. Temperature, white blood cell count with differential, and chills were monitored for each patient. In addition, the following were monitored for each infection site: (1) skin and skin structure: purulent drainage, induration, erythema, inflammation, pain, and ulceration; (2) lower respiratory tract: quantity and quality of sputum, productive cough, and chest roentgenogram; (3) urinary tract: frequency, dysuria, pyuria, and flank pain; (4) bacteremia: mental status and blood pressure; and (5) bone and joint: radiologic findings, erythrocyte sedimentation rate, purulent drainage, and localized pain, erythema, and inflammation. A n overall assessment of antibiotic treatment was made for each patient and rated as satisfactory, unsatisfactory, or indeterminate. The unsatisfactory rating included clinicalfailure as well as premature discontinuation of antibiotic secondary to an adverse event. T w o of the investigators reviewed each case to ensure consistency of evaluations. Statistical Analysis
Statistical calculations were performed by an independent group at the University of Minnesota using a Control Data Corporation Cyber 170/825 (Control Data Corporation, Bloomington, Minnesota) computer with S P S S software. S u m m a r y statistics (including means and standard deviations) were obtained for continuous and ordered variables, while frequency distributions were determined for categorical items. The two treatment groups were compared for baseline characteristics,
464
as well as antibiotic efficacy and safety. All categorical items were analyzed using chi-square with Yates' correction. Fisher's exact test was used when the sample sizewas lessthan 21. Variables that were not normally distributed underwent nonparametric analysis using the Mann-Whitney and KruskalWallis tests. Power calculations were performed to assess the potential for type II error. Unless otherwise noted, data are presented as mean ± standard deviation with the range in parentheses. All statistical analyses were conducted as two-tailed testswith p
