Nosocomial infections are frequent complications of hospitalization and also an important public health problem in developing countries, as well as in developed ones. The socioeconomic impact, ie, prolongation of hospitalization, mortality, and cost of these infections adversely effects patients and nations' economic well-being 1. The cost of nosocomial infections includes increased length of hospital stay, staff time, laboratory cultures of pathogens and antimicrobial treatment 234. Although, cost of antimicrobial treatment is an important part of health expenditure, data on this subject are extremely limited in Turkey.

The aim of our study was to determine daily antibiotic cost of nosocomial infection per infected patient in a university hospital.

Between January 1, 2000, and June 30, 2003, a total of 8460 patients were admitted to the adult ICUs. Overall, 817 patients developed 1407 episodes of nosocomial infections, accounting for an infection rate of 16,6%. Among them, 233 patients (mean age:50,1; sex ratio female:male 0,49) had only one nosocomial infection. Mean daily antibiotic cost was found $89,64 per infected patient ($8,56 to $359,28).

Among the sites of nosocomial infections, urinary tract infections had the lowest daily antibiotic cost per infected patient (Table 1). The mean daily antibiotic cost for pneumonia was the highest of all sites, but patients with bloodstream infection reached the highest range of daily cost ($31,31 to $359,28). In addition, mean daily antibiotic cost was found $162,35 for seventeen other infections including postoperative meningitis, mediastinitis, and empyema.

Out of 233 patients, 206 patients had microbiologically documented infections, 177 patients were infected by a single pathogen while 29 patients were diagnosed as having polymicrobial infections (Table 2). Pseudomonas aeruginosa was the most prevalent bacteria followed by Klebsiella spp. and Acinetobacter spp.. P. aeruginosa infections had the highest overall daily antibiotic cost per infected patient than other pathogens. Among 21 Staphylococcus aureus strains 15 (72%) were resistant to methicillin. The overall daily antibiotic cost of methicillin resistant S. aureus (MRSA) infections was two to three times higher than infections with susceptible strains. However, median daily cost per infected patient for MRSA infections were lower than susceptible strain infections.

Among the 350 antibiotic prescriptions for nosocomial infections, piperacillin-tazobactam and amikacin were the most prescribed antibiotics (Table 3). Carbapenems especially meropenem were the most expensive drugs.

Cost is an important factor which determines the physician's choice of medication to treat patients in spesific stiuations. In this study, we tried to demonstrate the daily cost of antimicrobial treatment of nosocomial infections according to site of infection, pathogen and antimicrobial agent.

In different studies, economical analysis regarding costs attributable to nosocomial infections has been evaluated and reported between $1018 to 2280 per infected patient 789. Jarvis et al reported that the estimated average costs of nosocomial infections were $558 to 593 for each urinary tract infection, $2734 for each surgical site infection, $3061 to 40000 for each bloodstream infection, and $4947 for each pneumonia 1.

Daily cost of antimicrobial treatment has been reported to be a significant extra cost attributable to nosocomial infections. In this study, we found an average daily antibiotic cost of $89,64 per nosocomial infection. It is clear that cost of overall antibiotic treatment for a period of approximately 10–15 days is $900 to $1350. Prolongation of hospital stay has been the major extra cost attributable to nosocomial infections in many reports 234, but in comparative case-control study from our country, Yalcin et al. 8 found that cost of antibiotic therapy of $1190 per infected patient, accounted for about 75% of the total extra cost. This finding may be due to the high prices of antibiotics in Turkey. To calculate the true costs of antibiotic therapy, hidden costs arising from intravenous administration, labor, serum antibiotic assay, monitoring hematological and biochemical indices and adverse effects of antibiotics must be considered 10. The present study does not include these relevant "hidden costs" that could substantially modify the total cost of an antibiotic treatment. Although, hidden costs were not calculated, an average daily antibiotic cost of a single nosocomial infection is found to be markedly high in our hospital.

This result is within the limits reported by other large economic studies, suggesting that our data is comparable to those found in other countries and with other assessment methods. In a French prevalence survey, Astagneau et al.11 reported an average daily antibiotic cost between FF 520 to 1085 (about $86 to $160) per nosocomial infection. French et al.12 and Haley et al.13 reported an average cost of antibiotic treatment of $190 and between $72 to $128 per nosocomial infection, respectively. In Turkey, Yalcin et al.14 found that daily antibiotic cost of nosocomial infections was $70 per patient.

The daily antibiotic cost varies markedly according to site of infection. Our study has demonstrated that pneumonia and bloodstream infections were associated with the highest daily antibiotic costs as reported in other studies 111314. Surgical site infections had also high daily antibiotic cost in our study. In their case-control study, Coello et al. reported that antibiotic therapy for surgical patients was the second most significant contributor to cost 15. In the present study, nosocomially infected patients that had only one nosocomial infection were considered for analysis. Clearly, antimicrobial treatment of patients with multiple nosocomial infections might be much more expensive.

P. aeruginosa infections had the highest daily antibiotic cost followed by other non-fermentative bacilli. Infections caused by P. aeruginosa are difficult to treat because of its virulence and relatively limited choice of effective antimicrobial agents, so, these infections often require combination therapy. Emergence of resistance in P. aeruginosa has been associated with increased morbidity, mortality, and costs 16. On the other hand, although the overall antibiotic cost of MRSA infections was higher than infections with susceptible strains, the daily antibiotic cost per infected patient with MRSA was lower with susceptible strain infections. MRSA infections are treated by glycopeptides which cost less than beta-lactams in our country. Astagneau et al reported that the daily antibiotic cost of multi-resistant bacterial infections such as multi-resistant P. aeruginosa infections, was 20% higher than susceptible infections, but the daily antibiotic cost per infected patient for MRSA infections was not higher than for susceptible strain infections 11.

Expensive antibiotics, such as piperacillin-tazobactam, carbapenems, cefepime, ciprofloxacin, teicoplanin were prescribed more commonly than the cheaper agents such as ampicillin-sulbactam, ceftriaxone or ofloxacin in our ICUs. These expensive antibiotics were mainly prescribed for resistant and severe gram-negative nosocomial infections, such as ventilator-associated pneumonia and postneurosurgical meningitis in ICU. Physicians may be forced to choose empirical antibiotic therapy with broad spectrum antimicrobials by increasing bacterial multi-resistance.

In conclusion, mean daily antibiotic cost was found $89,64 per nosocomial infection in our ICUs and nosocomial pneumonia had the highest daily antibiotic cost per infected patient. It is clear that cost of antibiotic therapy of nosocomial infections is an important part of extra cost attributable to nosocomial infection. Approximately one third of nosocomial infections are preventable by full implementation of the current infection control guideline recommendations 17. Each institution should develope empirical antibiotic guidelines according to its own local nosocomial infections data. Infection control measures, such as education of health care workers regarding antimicrobial agents and resistance; isolation of patients infected with multi-resistant organisms, should be implemented to reduce infections and expensive antibiotic prescriptions.

The author(s) declare that they have no competing interests.

DI collated and analyzed the data, participated in the study design and was principal writer of manuscript. ANY conceived the study. GO carried out the laboratory studies. RS, FG, OT and LM participated in the patient management. All authors read and approved the final manuscript.