Clinical Outcomes and Costs of Care of Antibiotic ... - CiteSeerX

2 downloads 0 Views 190KB Size Report
to July 2001, with clinically significant Pseudomonas aeruginosa isolates from ... Prevalence of resistant Pseudomonas aeruginosa infections was high at 39% ...
Clinical Outcomes and Costs of Care of Antibiotic Resistant Pseudomonas aeruginosa Infections Edwin L. Dimatatac, M.D., Marissa M. Alejandria, M.D., Cecilia Montalban, M.D., Cristina Pineda, R.M.T., Concepcion Ang, R.M.T. and Rachel Delino, M.S. ABSTRACT A prospective observational cohort study was conducted at a tertiary government teaching hospital to determine the prevalence of resistant isolates of Pseudomonas aeruginosa and its impact on the patients’ outcome, length of stay, cost of hospitalization. Adult patients admitted at the medical and surgical wards and intensive care units from February to July 2001, with clinically significant Pseudomonas aeruginosa isolates from clinical specimens and hospital stay of at least three days were included in the study. All patients were followed up from the time of detection of the baseline Pseudomonas aeruginosa isolate until discharge or death. Cultures were repeated from the same site of the initial isolate after 72 hours of antibiotic administration. Exposure variables obtained were severity of illness score, co-morbid illness, foci and source of infection, presence of invasive devices, surgical intervention and history of antibiotic use. Microbiologic outcome were categorized as susceptible, baseline resistance, persistence and colonization of Pseudomonas aeruginosa; clinical outcome included mortality or survival, development of superinfection, secondary bacteremia, sepsis/septic shock. To estimate resource utilization and costs of care, the following variables were collected: length of hospital stay following the detection of the baseline isolate and daily antibiotic and antibiotic administration costs following detection of the baseline isolate. Two hundred twenty one patients had clinically significant isolates out of 310 P. aeruginosa isolates identified at the laboratory. Evaluable patients were 199, of whom 47% had susceptible isolates, 17% had emergence of resistance, 16% had baseline resistance, 14% were assessed as colonizers and 6% were assessed to have persistence of Pseudomonas aeruginosa infection. Most of the infections were hospital acquired (86%) for both resistant and susceptible isolates. Emergence of resistance was highest in the medical wards (50%). Severity of illness scores (p = 0.003) and mortality rates (p < 0.001) were significantly higher among those with resistant isolates compared to those with susceptible isolates. Patients with emergence of resistance had significantly longer median duration of hospital stay than those with susceptible isolates (p < 0.001). Median antibiotic and antibiotic administration costs did not differ among those with resistant and susceptible isolates (p=0.11). Development of resistance was highest with cefpirome, ciprofloxacin and ceftazidime; and lowest with the carbapenems and cefepime. Prevalence of resistant Pseudomonas aeruginosa infections was high at 39% with an overall mortality of 41%. Patients stayed significantly longer in the hospital with emergence of resistance. Infection control policies and guidelines on rational antibiotic use need to be reviewed and intensified to minimize costs of care. [Phil J Microbiol Infect Dis 2003; 32(4):159-167] Key Words: Pseudomonas aeruginosa, antibiotic-resistance, sepsis, septic shock

INTRODUCTION Pseudomonas aeruginosa is known for its ability to resist killing by a variety of antibiotics. The minimal nutritional requirements of Pseudomonas, as evidenced by its ability to grow in distilled water and its tolerance to a wide variety of physical conditions, contribute to its ecologic success and ultimately to its role as an effective opportunistic pathogen. Pseudomonas aeruginosa is primarily a nosocomial pathogen. In the annual surveillance of nosocomial infections by the Centers for Disease Control and Prevention from 1990 to 1996, it is the 2nd most common etiology of nosocomial pneumonias; 3rd for

urinary tract infections and 4th for surgical site infections.1 Likewise in a hospital-wide surveillance of nosocomial infections conducted by the Infection Control Committee of the Philippine General Hospital in 1989, Pseudomonas aeruginosa was the most common organism isolated from all sites of infection (37%).2 Resistance to antimicrobial agents is an increasing clinical problem and is a recognized public health threat. Pseudomonas aeruginosa has a particular propensity for the development of resistance. It is naturally resistant to many antibiotics because of its relatively impermeable outer membrane and it can also easily acquire resistance, creating challenging therapeutic scenarios. All known mechanisms of β-lactam resistance can be found in this specie namely: β-lactamase production, altered outer-membrane permeability, active efflux and altered penicillin-binding proteins.3 Thus, infections due to this organism are difficult to treat because of the possible coexistence of several mechanisms of resistance in the same strain; its capacity to produce a variety of virulence factors and the relatively limited choice of effective anti-pseudomonal antibiotics. Furthermore, emergence of resistance during therapy with these agents has been recognized as a cause of treatment failure. The reason that antibiotic resistance leads to adverse outcomes is due to the increased likelihood that the patient will receive ineffective or suboptimal antibiotic therapy. The development of resistance to all available antibiotics in some organ-isms then precludes the effectiveness of any antibiotic regimen. Organisms that are resistant to all known effective antimicrobials pose a serious threat to hospitalized patients. Thus far, two gram-positive organ-isms have been described as being resistant to all antibiotics: vancomycin-resistant enterococci and vancomycin-resistant, methicillin-resistant S. aureus. Similarly, gram-negative bacteria may also become resistant to all available antiThis is most likely to occur in an organism, such as Pseudomonas aeruginosa in which resistance to different classes of antibiotics has already been described. Comparative analyses of the frequency of emergence of resistance associated with the different classes of antipseudomonal drugs are scarce. In a cohort study of 271 patients conducted from 1994 to 1996 in a tertiary-care teaching hospital in Boston, Massachusetts, Pseudomonas aeruginosa resistant to at least one of the antipseudomonal study agents emerged in 28 patients (10.2%) giving an incidence of 7.4 patients per 1,000 patient days.4 Interestingly, this study reported that the use of ceftazidime was associated with the lowest risk for the emergence of resistance, while imipenem had the highest risk. The study also analyzed the hazard ratios for emergence of resistance to each individual antipseudomonal agent associated with treatment with the same agent. Similarly, ceftazidime had the lowest hazard ratio at 0.8 followed by piperacillin at 5.2, ciprofloxacin at 9.2, with imipenem having the highest hazard ratio of 44.4 The emergence of resistance in Pseudomonas aeruginosa limits therapeutic choices and is associated with increased rates of morbidity and mortality, higher costs and prolonged hospitalization relative to antibiotic-susceptible bacteria. In a similar study in Boston as cited above, emergence of resistance of Pseudomonas aeruginosa isolates was associated with a 3-fold increase in mortality, a 9-fold higher rate of secondary bacteremia, and a 2.1 increase in hospital days, and higher hospital charges.5 Further review of the characteristics of these multi-resistant isolates revealed that in 16 of the 22 cases (72%), isolation of susceptible P. aeruginosa preceded the culture of the multiresistant isolate. The overall mean length of stay after isolation of multi-resistant P.

aeruginosa was 18 days, and the mean cost of the admission during which the first multiresistant isolate was cultured was $54,081, compared with a mean of $22,116 for admission of patients with infection with susceptible P. aeruginosa. Indiscriminate use of antibiotics leads to the development of resistance of initially sensitive strains of organisms, and possible destruction of the normal microbial flora. Local studies that quantitatively examine the health and cost impact of resistant organisms in our setting are lacking. Information on the emergence of resistance with different antibiotics can be of practical use in guiding empiric therapeutic choices. This prospective observational study aims to compare morbidity, mortality and costs of health care associated with infections due to antimicrobial resistant and susceptible strains of clinically significant P. aeruginosa. Baseline data derived from this study can be used as bases for instituting preventive measures and formulating recommendations on rational antibiotic use. MATERIALS AND METHODS All adult patients, 18 years old and above, admitted at the adult medical and surgical wards and intensive care units of the Philippine General Hospital from February 1, 2001 to July 31, 2001 who fulfilled the following criteria and consented to join the study comprised the study population: 1. Isolation of clinically significant P. aeruginosa from clinical specimens 2. Susceptibility of the first P aeruginosa isolate to at least one of the antipseudomonal drugs 3. Treatment with at least one of the anti-pseudomonal drugs based on susceptibility results 4. Hospital stay of at least 3 days Excluded were patients whose initial P. aeruginosa isolates were resistant to all anti-pseudomonal antibiotics, patients with poly-microbial isolates and those with P. aeruginosa isolates from another hospital. Study Procedure All P. aeruginosa isolated from all clinical specimens submitted to the UP-PGH microbiology laboratory were collected. Antibiotic susceptibility testing of all clinically significant isolates was performed at the UP-PGH Infectious Disease Laboratory, using the Kirby and Bauer method. The National Committee for Clinical Laboratory Standards guidelines was followed. All antipseudomonal drugs approved for use by the Therapeutics Committee of the Philippine General Hospital were tested namely: ceftazidime, ciprofloxacin, imipenem, meropenem, piperacillin-tazobactam, amikacin, cefepime, cefpirome. Isolates with intermediate susceptibility were considered resistant. Patients included in the study received antipseudomonal antibiotics based on susceptibility testing. The initial Pseudomonas aeruginosa isolate of each patient was considered the baseline isolate. Patients were examined and the following information were collected:

demographic data, co-morbid illnesses, focus/foci of infection, source of infection, presence of invasive devices, presence of surgical intervention, previous and current antibiotic use. Severity of illness was scored using the PGH Mortality Prediction Model.6 This is a bedside scoring system designed and validated in the PGH medical intensive care unit and in sepsis patients in the ICUs and wards.7 Repeat cultures from the same site of the initial isolate were performed after 72 hours of antibiotic administration to assess their microbiologic outcomes. Repeat cultures of other sites were done together with other laboratory examinations, as deemed necessary depending on the clinical response or detection of other sites of infection. Clinical response was assessed based on fever pattern, physical examination findings, CBC and other pertinent laboratory results. Outcome measures All patients were prospectively followed up from the date of detection of the baseline isolate until discharge or death. Microbiologic outcomes were categorized as follows: baseline resistance, emergence of resistance, persistence, colonization and susceptible. The following clinical events were determined during hospitalization: mortality or survival, development of superinfection, secondary bacteremia, sepsis/septic shock. To estimate resource utilization and costs of care, the following variables were collected: length of hospital stay following the detection of the baseline isolate and daily antibiotic and antibiotic administration costs following detection of the baseline isolate. The delay in effective treatment related to resistance was calculated as the number of days that a patient is treated with an anti-pseudomonal agent to which the organism is resistant and simultaneously not receiving an antibiotic active against the baseline isolate. Whereas, attributable cost of resistance was defined as incremental costs of care for an infection due to a resistant isolate minus the costs of care of infection with a susceptible strain of the same organism.8 Operational Definition of Terms Clinically significant isolate is defined as an isolate that is responsible for the ongoing infection as supported by the history, physical examination findings and other laboratory results. Susceptible Pseudomonas aeruginosa isolate is defined as susceptibility of the baseline isolate to the current drug and there is clinical improvement or eradication of the isolate on subsequent culture. Treatment with the current drug is continued as deemed necessary. Baseline resistance is defined as resistance of the baseline isolate to the currently administered drug and this drug is shifted to at least one of the antipseudomonal drugs based on susceptibility patterns. Emergence of resistance is defined as the subsequent detection of Pseudomonas aeruginosa that is resistant to the current drug and there is no clinical improvement. This drug is shifted to at least one of the antipseudomonal drugs based on susceptibility patterns. Persistence of Pseudomonas aeruginosa is defined as the subsequent detection of Pseudomonas aeruginosa that is sensitive to the current drug but there is no clinical

improvement and the drug is shifted to at least one of the anti-pseudomonal agents based on the susceptibility pattern. Colonization of Pseudomonas aeruginosa is defined as the detection of Pseudomonas aeruginosa on subsequent culture but there is clinical improvement and the current antibiotic is continued as deemed necessary. Secondary bacteremia is defined as a positive blood culture for Pseudomonas aeruginosa at the same time as or after the detection of baseline isolates from respiratory, urine, exudates or other specimens. Superinfection is defined as an infection caused by any organism other than P. aeruginosa that was isolated on subsequent culture. Sepsis is defined as clinical evidence of infection, plus evidence of a systemic response to infection as manifested by two or more of the following conditions: Temperature of >38oC or 90 beats/minute; Respiratory rate >20 cycles/minute or PaCO2 12 cells/mm3 or 10% immature band form.9 Septic shock is defined as sepsis with hypotension despite adequate fluid resuscitation along with the presence of perfusion abnormalities that may include, but not limited to, lactic acidosis, oliguria, or acute alteration in mental status.9 Sample Size To estimate the prevalence of resistant Pseudomonas aeruginosa isolates with a precision of + 5% and a 95% level of significance, 191 Pseudomonas aeruginosa isolates are needed. This is based on the resistance patterns of P. aeruginosa isolates from the medical wards and intensive care unit, with ciprofloxacin, having the highest resistance rate of 42.7% for the year 2000. Data Analysis Descriptive statistics was used to describe the demographic data and outcome measures. For the analysis of microbiologic outcomes, the categories of baseline resistance, emergence of resistance and persistence were classified as resistant isolates, while the categories of susceptible and colonization were classified as susceptible isolates. 1. Susceptible - Patients treated with at least one of the anti-pseudomonal drugs in which the baseline isolate is susceptible to the current drug and there is clinical improvement and eradication of the isolate on repeat culture (susceptible P. aeruginosa) and those patients who still had positive culture for P. aeruginosa on subsequent isolate but there is clinical improvement (colonization). 2. Resistant - Patients treated with at least one of the anti-pseudomonal drugs in which the baseline isolate is resistant to the current drug (baseline resistance) and those patients who still had positive cultures for P. aeruginosa on subsequent culture and there is no clinical improvement (emergence of resistance and persistence of P. aeruginosa).

For the cost analysis, the length of hospital stay, antibiotic and antibiotic administration expenditures for each patient were computed based on the PGH 2001 price list. To test for significant differences among the microbiologic outcomes, KruskalWallis was used. Level of significance was set at p < 0.05. All data were entered in Microsoft EXCEL and Epi Info 6.4 was used for statistical analysis. RESULTS Out of 310 patients from the medical and surgical wards and ICUs with Pseudomonas aeruginosa isolates identified at the PGH microbiology laboratory during the 6-month period, 221 patients with clinically significant isolates were included in the study. Twenty-two patients were unevaluable for the microbiologic outcome because they died within 48-72 hours after isolation of the baseline organism. Excluded were two patients with isolates resistant to all anti-pseudomonal antibiotics, four patients 3 *In Philippine Pesos

Persistence

Colonization

Susceptible

N=31 19 [14-37] 5,392.32 [3067.8315655] 6,015.20

Emergence of Resistance N=34 35.5 [30-50] 11,853.65 [413624705.64] 12,234.45

N=13 14 [8-28] 6,134.42 [34722671.60] 6,757.32

N=27 36 [27-49] 6,196.00 [250638389.05] 7,169.39

N=94 26.5 [23-30] 4,026.38 [2479.507695] 4,606.17

[339.7816005.95] 17

[4476.95271101] 19

[61921940.83] 12

[3082.4539204.80] 14

[2836.957966.95] 32

(56%)

(92%)

(52%)

(34%)

(55%)

p-value Kruskal Wallis p=0.001 p=0.112 p=0.07

p=0.003

Legend: Caz = ceftazidime, Cip= ciprofloxacin, Amik =amikacin, Tzp = piperacillin-tazobactam, Imp = imipenem, Mem=meropenem, Fep=cefepime, Cpo=cefpirome Figure 1. Per cent resistance of Pseudomonas aeruginosa isolates, Philippine General Hospital, Feb-July 2001, N=311

DISCUSSION Pseudomonas aeruginosa continues to be a leading cause of serious infections, particularly nosocomial infections as shown in this study. This study analyzed 221 patients at the Philippine General Hospital from whom clinically significant Pseudomonas aeruginosa was isolated. Prevalence of resistant Pseudomonas infection was particularly high at 39% in this institution. Majority of the cases were nosocomial and characterized by prolonged and severe infections. Severity of illness score was highest among patients with persistent Pseudomonas aeruginosa isolates. Consequently, mortality was high especially among those with resistant and persistent isolates, emphasizing the difficulty in treating these organisms once resistance has emerged. Overall mortality was 41%, 10% of whom died within 48-72 hours after isolation of the organism. Not surprisingly, the patients with resistant and persistent isolates died due to the infection primarily. Baseline resistance resulted in a delay of administration of effective antibiotic treatment of about 48 hours. Development of resistance was highest with ciprofloxacin and ceftazidime, while it was lowest with the carbapenems and cefepime. This finding is in contrast with that of Carmeli et al in Boston wherein ceftazidime was associated with the lowest risk for emergence of resistance (adjusted hazard ratio 0.7), while imipenem had the highest risk (adjusted hazard ratio 2.8).4 These contrasting results reflect the differences in patterns of antibiotic usage. In the Philippine General Hospital, physicians frequently use ciprofloxacin and ceftazidime as initial empiric antibiotic choices for suspected pseudomonal infections. Emergence of resistance was highest in the medical wards, where understandably antibiotic exposure and usage was highest compared to the other wards. Resistance patterns of clinically significant isolates obtained in this study were higher compared to the susceptibility pattern of 789 P. aeruginosa laboratory isolates from clinical specimens of patients at the Philippine General Hospital for the year 2001, without taking into consideration clinical significance. The resistance rates of the laboratory isolates were as follows: amikacin (3%); cefepime (3%); ceftazidime (12%); meropenem (14%); imipenem (19%); piperacillin-tazobactam (22%); ciprofloxacin (25%).10 The trend however, is similar with ciprofloxacin consistently with the highest resistance rate. This emphasizes the importance of clinical correlation in interpreting antibiotic susceptibility test results. Furthermore, the emergence of antibiotic resistance in a previously susceptible strain makes these infections even more difficult to cure. Resistance emerged during treatment with each class of antibiotic and did not appear to be prevented by the use of combination therapy with aminoglycosides. Emergence of resistance leads to adverse outcomes by rendering existing antibiotic therapy microbiologically ineffective, thus causing progressive or relapsing clinical infection. Multi-resistant Pseudomonas aeruginosa evolves in a stepwise manner in patients with difficult to treat infections upon exposure to different classes of anti-pseudomonal antibiotics. Reversion to susceptibility to one or more antibiotic agents within the same admission can occur. This phenomenon maybe explained by a dynamic situation in which numerous subpopulations of the same infecting strain are present at any point in time.

Analysis of resource use was limited to duration of hospital stay, antibiotic and antibiotic administration costs only. Results of this partial cost analysis showed that emergence of resistance significantly prolonged the hospital stay of patients but it did not affect the antibiotic and antibiotic administration costs per se. Since this is a limited analysis of costs, a number of direct and indirect costs and effects of emergence of antibiotic resistance are not accounted for in this study. Indirectly, prolonged hospital stay will increase fixed costs such as hospital charges (cost per day per bed) and indirect costs, which include opportunity costs due to loss of income. Other costs and effects to be considered are laboratory costs for screening and diagnosis, cost of patient isolation, physician, nursing and infection control staff time and occurrence of other infections and complications. An observational study such as this has a number of limitations, including the potential for confounding due to lack of randomization and the use of more than one type of antibiotics for some patients. Because of the small sample size, logistic regression analysis to adjust for confounders was not performed. Another limitation is that, we did not do typing of resistance patterns using pulsed-field gel electrophoresis of all available pairs of isolates (before and after resistance emerged) to confirm that resistance emerged in a susceptible isolate. Determination of minimum inhibitory concentrations for each antibiotic was also not done. From a practical viewpoint, physicians should always anticipate the potential for emergence of resistance when managing patients with Pseudomonas infections. Routine culture and susceptibility testing should be done as soon as possible to detect emergence of resistance once signs of treatment failure is observed. CONCLUSION Overall prevalence of resistant Pseudomonas aeruginosa infections was high at 39% in this institution with an overall mortality of 41%. Patients with emergence of resistance had significantly longer duration of hospital stay. Patterns of antibiotic usage need to be reviewed and guidelines on rational antibiotic use need to be reinforced. Infection control measures must be intensified to minimize costs of care. Acknowledgements We gratefully acknowledge funding from the Philippine College of Physicians and the following pharmaceutical companies: ASTRA ZENECA, Wyeth, Glaxo Smith Kline, United American Pharmaceutical, Bayer and Merck Sharp and Dohme. We also express gratitude to Winnie Castillo, RN for encoding the data, Mr. Mark Javellosa for statistical advice, the fellows and staff of the UP-PGH Section of Infectious Diseases for their invaluable assistance and support; and Myrna Mendoza, MD for expert advice during the conduct of this study.

REFERENCES 1. 2. 3. 4. 5.

National Nosocomial Infection Surveillance (NSSI) report: data summary from October 1986-April 1996. Am J Infect Control 1996; 24(5):380-388. ICC Bulletin. Jan-May 1990; 111(1). Pechere JC, Kohler T. Patterns and modes of β-lactam resistance in Pseudomonas aeruginosa. Clin Microbiol Infect 1999; 5(suppl 1):S15-S18. Carmeli Y, Troillet N, Eliopoulos GM, Samore MH. Emergence of antibiotic-resistant Pseudomonas aeruginosa: comparison of risks associated with different antipseudomonal agents. Antimicrob Agents Chemother 1999; 43(6):1379-1382. Carmeli Y, Troillet N, Karchmer AW, Samore MH. Health and economic outcomes of antibiotic resistance in Pseudomonas aeruginosa. Arch Intern Med 1999; 159(10):1127-1132

6. 7. 8. 9. 10.

Cordero CP, Ngelangel C, Beltran A, et al. Derivation of a mortality prediction model at the intensive care unit of the Philippine General Hospital. Phil J Internal Medicine 1994; 32:275-279. Fonbuena EG, Fadreguilan E, Timbreza F, Alejandria M, Lansang M. Comparison of the predictive accuracy of two scoring systems in medical patients with sepsis at the Philippine General Hospital. Phil J Microbiol Infect Dis 2002; 31(3):95-100. Howard D, McGowan J, Packard R, Scott RD, Solomon SL. Measuring the economic costs of antimicrobial resistance in hospital settings: summary of the Centers for Disease Control and Prevention-Emory Workshop. Clin Infect Dis 2001; 33(9):1573-1578. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical care Medicine. Chest 1992; 101(6):1644-1655. Antimicrobial Resistance Surveillance of Philippine General Hospital. Microbiology Section, Department of Laboratories, 2001.

APPENDIX PGH-SIMPLIFIED MORTALITY PREDICTION MODEL Parameters Overt GI bleeding Coma/deep stupor Needs mechanical ventilator CPR within 24 hours prior to admission Age >45 SBP 22/min Total score of >3 indicates poor prognosis

Assigned Score 5 4 3 2 1 1 1

Yes

No

Patient’s Score