Hospital costs of nosocomial multi-drug resistant Pseudomonas aeruginosa acquisition

Antibiotic resistance is a growing clinical problem and a major public health threat. Infections caused by antibiotic resistant microorganisms usually result in significantly higher morbidity, longer hospitalization, increased mortality rates, and excess health care costs compared with antibiotic-susceptible microorganisms [1, 2]. Although much attention has focused on emerging antibiotic-resistant Gram-positive bacteria such as methicillin-resistant Staphylococcus aureus and vancomycin-resistant Enterococcus, resistance within Gram-negative bacilli continues to rise [3], occasionally creating situations in which few or no antibiotics that retain activity are available.

Pseudomonas aeruginosa is a Gram-negative bacterial pathogen that causes severe nosocomial infections [4]. This microorganism is a leading cause of nosocomial infections and is responsible for 10% of all hospital-acquired infections, ranking second among Gram-negative pathogens [5, 6]. Multi-drug resistance to antipseudomonal antibiotics is a common and increasing problem in some hospitals [7, 8]. Infections by multi-drug resistant P. aeruginosa (MDRPA) are associated with increased morbidity [9], mortality [7, 10], and economic impact [11]. A number of studies have evaluated the impact of P. aeruginosa antibiotic resistance through assessment of in-hospital mortality rates and length of hospitalization [7–11]. However, fewer studies have quantitatively examined in detail the hospital economic impact of MDRPA infections and most have been limited to case series studies and outbreaks [11, 12].

Not only morbidity and mortality, but also the economic cost may be a sensitive measure to quantify the burden of antimicrobial resistance [13]. In addition, estimating cost outcomes due to antimicrobial resistance may be useful to prompt hospitals and health care providers to introduce and support initiatives to prevent such infections, to influence health care providers to follow isolation guidelines, and to make appropriate use of antibiotics. Moreover, data can guide policy makers who take decisions on the funding of programs to track and prevent the spread of antimicrobial-resistant microorganisms.

The objective of this study was to assess the hospital economic costs of MDRPA acquisition in a tertiary hospital in Barcelona (Spain) during the period 2005–2006.

Overall, 402 incident P. aeruginosa positive cultures among hospitalized patients were identified during the study period. Their distribution by antibiotic susceptibility pattern was 149 (37.1%) non-resistant, 119 (29.6%) resistant, and 134 (33.3%) multi-drug resistant.

We assessed the hospital economic impact of nosocomial MDRPA acquisition in a tertiary hospital. Our comparison group was composed of patients with positive cultures admitted. Active surveillance of nosocomial infections was not implemented in our centre, thus most of the cases included in the present study corresponded with infected patients. Therefore, the present results should be interpreted as the impact of antibiotic resistance on hospital costs. The mean hospital total cost per admission was Є12,351 for patients who acquired resistant and €15,265 for MDRPA strains, which was substantially higher than the Є4,933 for those with non-resistant strains. On multivariate analysis, we found an independent contribution of P. aeruginosa multi-drug resistance acquisition to increased total cost per hospital admission, which was more than 1.7 fold higher in comparison with non-resistant microorganisms. Pharmacy cost showed the strongest increase, which was higher than 3-fold.

Antibiotic multi-resistant infections caused by Gram-negative bacilli are emerging in hospital settings and are becoming a major clinical and public health issue. Despite of the absence of an international consensus on the definition of multi-drug resistance in P. aeruginosa complicating comparison between studies [21], we found a higher prevalence of multi-drug resistance than previous reports from the European Antimicrobial Resistance Surveillance System (EARSS) Annual Report for 2006 http://www.rivm.nl/earss/result/Monitoring_reports/, 18% of P. aeruginosa isolates were found to be multidrug-resistant, i.e. resistant to three or more antibiotics from the EARSS protocol [5]. This change in trends strengthens the importance of assessing the impact of the increase in multi-drug resistance on hospital outcomes not only in terms of mortality or prolongation of hospital stay, but also on economic costs.

Globally, it is estimated that infections due to antimicrobial-resistant organisms have higher costs ranging between $6,000 and $30,000 than do infections due to antimicrobial susceptible organisms [13]. In our study, the mean cost per hospital admission with a positive culture of MDRPA was €15,265, which was substantially higher than that of hospital admissions infected with non-resistant strains (mean of €4,933). Previous studies have estimated the hospital cost increase due to MDRPA infections. In an ICU outbreak, Bou et al. [12] reported that the emergence of antibiotic multi-resistance in P. aeruginosa infections was significantly associated with a high median economic cost (€72,356 vs €17,090; P < 0.004). Lautenbach et al. [22] reported that patients infected with imipenem-resistant Pseudomonas aeruginosa had greater hospital costs (US$81,330 vs. US$48,381; P < 0.001). In a matched cohort study Carmeli et al. [9] estimated that emergence of resistance to P. aeruginosa increased the total hospital charges by $7,340. In a case-series study, Harris et al. [11] reported that the mean cost of the admission during which the first multi-resistant isolate of P. aeruginosa was cultured was of $54,081, compared with a mean cost of $22,116 for patients infected with susceptible strains. Most of these studies were limited to case series and outbreaks among patients in ICU, which may partly explain the differences in cost estimations in comparison with current estimations. Also the inclusion in our study of both infections and colonizations could result in underestimated costs. In addition, diverse cost estimation methodology could influence differences in cost estimations when comparing studies.

When analyzing different types of hospital cost, we found a remarkable increment in hospital cost due to MDRPA strains in pharmacy cost. Although not exclusively, pharmacy excess costs associated with multi-drug resistant microorganisms may partly be due to delayed appropriate antibiotic therapy and the need to use more expensive antibiotics, which implies a substantial increase in drug therapy cost. In our experience, colistin (polimixin-E) is the first choice antimicrobial for MDRPA infections [14]. Comparatively with other antipseudomonals colistin is as expensive as other broad- spectrum anti-pseudomonas antibiotics such as carbapenems or the pip/tazo combination. Importantly, these excess costs are potentially avoidable.

Notable among the strengths of the study is that information was available from almost all P. aeruginosa acquisitions detected among hospital admissions during the study period. This study has some limitations. To estimate the effect of antibiotic resistance on the economic cost of nosocomial acquisitions, patients with antibiotic resistant microorganisms were compared with those without resistant organisms for hospital cost outcomes. The problem with this simple comparison is that sicker patients with longer length of hospital stay, greater severity, and comorbidities are more likely to acquire resistant organisms, leading to substantial confounding bias. However, we addressed this problem by detailed adjustment to control as much as possible the factors other than antibiotic resistance that contributed to adverse outcomes, such as severity, and UCI stay post infection. This enabled us to isolate the independent effect of P. aeruginosa acquisition on hospital costs. Another relevant limitation in estimating hospital cost in relation to infection acquisition involves time and length bias [23]. However, no statistical solution exists for models involving a quantitative dependent variable such as cost, which cannot be directly related to time. Thus, to address this methodological issue we used costs after infection as dependent variable, which can reduce cost overestimation due to time and length bias. On the other hand, using generalized linear models (GLM) to estimate hospital costs could result in biased and over-estimated costs. However, a recent study comparing different analytic approaches did not demonstrate that multistate methods are better than GLM or propensity score adjustments [24]; and the most important issue in calculating the cost or the economic impact relies on using the real cost for each day of stay and for each patient and on the quality of the data [24]. Moreover, our cost estimates most closely represent the direct hospital perspective, which provides a partial view of the health care impact of antibiotic resistance. Thus, we believe that the hospital perspective is to some extend also underestimated because we did not account for extra costs to the institution at large attending patients with MDRPA infections, such as surveillance cultures, isolation supplies, and loss of beds due to the need to isolate a patient. And in addition, the effects extending beyond hospitalization were underestimated by this study. Finally, data from one centre and one particular health care system could limit broader applicability of the present results.

In conclusion, MDRPA acquisition is becoming a serious issue in hospital settings and is independently associated with a substantial increase in medical costs. Optimal use of existing antimicrobial agents through education of health professionals, antibiotic policies (including antibiotic stewardship), implementation of infection control measures (e.g., hand hygiene, environmental cleaning, active screening cultures and isolation), and reducing the length of patient stay should be the strategies aimed at preventing the emergence and spread of antibiotic resistance in order to limit their adverse consequences, including their economic cost.