Attributable costs of ventilator-associated lower respiratory tract infection (LRTI) acquired on intensive care units: a retrospectively matched cohort study
© Leistner et al.; licensee BioMed Central Ltd. 2013
Received: 16 November 2012
Accepted: 29 March 2013
Published: 4 April 2013
Lower respiratory tract infections (LRTI) are the most common hospital-acquired infections on ICUs. They have not only an impact on each patient’s individual health but also result in a considerable financial burden for the healthcare system. Our aim was to determine the costs and the length of stay of patients with ICU-acquired LRTI.
We used a retrospectively matched cohort design, comparing patients with ICU-acquired LRTI and ICU patients without LRTI. LRTI was diagnosed using the definitions of the Centers for Disease Control and Prevention (CDC). Study period was from January to December 2010 analyzing patients from 10 different ICUs (medical, surgical, interdisciplinary). The device utilization ratio was defined as number of ventilator days divided by number of patient days and the device-associated LRTI rate was defined as number of ventilator associated LRTI divided by number of ventilator days. Patients were matched by age, sex, and prospectively obtained Simplified Acute Physiology Score II (SAPS II). The length of ICU stay of control patients needed to be at least as long as that of LRTI-patients before onset of LRTI. We used the Wilcoxon signed-rank test for continuous variables and the McNemar’s test for categorical variables.
The analyzed ICUs had 40,772 patient days in the study period with a median ventilation utilization ratio of 56 (IQR 42–65). The median device-associated LRTI rate was 3.35 (IQR 0.96-5.36) per 1,000 ventilation days. We analyzed 49 patients with ICU-acquired LRTI and 49 respective controls without LRTI. The median hospital costs for LRTI patients were significantly higher than for patients without LRTI (45,041 € vs. 26,467 €; p < .001). The attributable costs per LRTI patient were 17,015 € (p < .001). Patients with ICU acquired LRTI stayed longer in the hospital than patients without (36 days vs. 24 days; p = 0.011). An LRTI lead to an attributable increase in length of stay by 9 days (p = 0.011).
ICU-acquired LRTI is associated with increased hospital costs and prolonged hospital stay. Hospital management should therefore implement control measurements to keep the incidence of ICU-acquired LRTI as low as possible.
Lower respiratory tract infections (LRTI) are the most frequent infections acquired on intensive care units (ICU) [1–3]. However, the incidence among ICU patients varies among different studies and depends on the definition of LRTI and the method of surveillance [2, 3]. The economic impact of hospital-acquired LRTIs on healthcare systems is meanwhile an acknowledged topic in literature [2–4]. Nevertheless, there are few clinical studies assessing the attributable costs of ICU-acquired LRTIs in a DRG-based (diagnosis related groups) healthcare system [5–8]. Those studies analyzed only data from the USA. Resource utilization in the European and American healthcare systems are only partially comparable. It is unclear how those differences influence the results of economic studies. This substantiates the need for European studies. To our knowledge, this is the first study on LRTI costs in a DRG-system from Europe.
This study was conducted at the Charité University Medical Center, a 3,213 bed tertiary care university hospital. Approximately 136,000 patients are admitted each year and approximately 560,000 patients are treated as outpatients. The average stay in our hospital is 6.65 days and the rate of bed utilization is 83.6%. The infection control department of the Charité performs routinely surveillance for nosocomial LRTI on 10 different ICUs (medical, surgical and interdisciplinary) using the method of the German hospital infection surveillance system for intensive care units (ITS-KISS) [9, 10]. ITS-KISS uses the definitions for hospital acquired infections from the Centers for Disease Control and Prevention (CDC) .
Study design and data collection
We used a retrospectively matched cohort design to study costs and outcome of patients with LRTI acquired on one of our intensive care units (ICU). The study period was from January 1st 2010 to December 31st 2010. In the following, patients who acquired an LRTI while their stay on an ICU are classified as cases and patients without LRTI are classified as controls. Cases and controls were prospectively found by trained infection control nurses performing the surveillance on our ICUs for ITS-KISS. The device utilization ratio (number of ventilator days divided by number of patient days on the analyzed ICUs) and the device-associated LRTI rate (number of ventilator associated LRTIs divided by number of ventilator days on the analyzed ICUs) were calculated. Additional information for each patient was gathered by hospital file search. For all patients included in our study the following characteristics were collected: Age, sex, the Simplified Acute Physiology Score (SAPS) at entry in the ICU, the Charlson Comorbidity Index (CCI), overall length of stay, length of stay before onset of LRTI, length of stay after onset of LRTI, length of stay on the ICU and ventilation hours. The study based on secondary clinical datasets that did not include identifiable information. Therefore, further ethics statement or informed consent was not required.
Costs and outcome
Data on hospital costs derived from true hospital costs (hospital charges) and were provided by the financial control department of the Charité University Medical Center. The analyzed costs cover the direct costs due to treatment and diagnostics and the indirect costs due to activities without patient contact (e.g. administration, hospital maintenance). The individual case charges were estimated based on definite performances and on settlement keys (e.g. nurse working time per patient). Costs were broken down into costs for medical staff, nursing staff, assistant medical technicians, medical products and for pharmacy. Daily costs were obtained by total hospital costs divided by number of hospital days. Reimbursement per patient was calculated on the basis of the diagnosis related groups (DRG), provided by the financial control department. The attributable costs and attributable length of stay (LOS) of hospital acquired LRTI were calculated as median of the differences in total costs of cases and their respective controls. The patient case weight serves as a weighing factor to give medical cases an economic dimension to demonstrate their economic severity. The case weights are estimated on the basis of several parameters: e.g. the patients’ primary and secondary diagnoses, the therapy, age, sex, ventilation time.
Cases and definitions
Cases were defined as patients who were 18 years or older at the time of hospital admission and who acquired a LRTI or a pneumonia on one of the analyzed ICUs. Cases of pneumonia and LRTI were subsumed as LRTI cases. Cases of LRTI were diagnosed using the CDC definitions and were considered ICU-acquired if no evidence that the infection was present or incubating at the time of admission to the ICU . The used CDC definitions include clinically defined pneumonia, laboratory confirmed pneumonia and lower respiratory tract infection other than pneumonia. A laboratory confirmation is required for the latter two definitions. The included patients had to be admitted and discharged within 2010. Patients’ whose stay ended later than 2010 were excluded. Mortality was defined as hospital mortality.
Controls and matching criteria
All patients who were discharged in 2010 and who were admitted to an ICU in 2010 were considered a potential control. Patients who were younger than 18 years at the time of admission and patients who stayed less than the minimum of 3 days on the ICU were excluded. Control patients who were clinically diagnosed with a LRTI or who showed signs of LRTI in the retrospective file search were excluded as controls. Cases and controls were matched in a ratio of 1:1, using following criteria: Age (± 5 years), sex, and SAPS (± 10 points) and the total length of ICU stay of controls needed to be at least as long as that of cases before onset of LRTI.
For the case and control patients, we calculated the median and interquartile range (IQR) for continuous parameters and number and percentage for categorial parameters. For evaluation of the matching criteria application and to test differences between groups, we used the Wilcoxon signed-rank test for continuous variables and the McNemar’s test for binary variables. Data were analyzed using PASW Statistics 18 (SPSS Inc., 2009, Chicago, Illinois). P ≤ 0.05 was considered significant.
Characteristics of the analyzed ICUs and patients
Descriptive characteristics of cases with ICU-acquired LRTI and control patients without LRTI
Cases (N = 49)
Controls (N = 49)
Time with mechanical ventilation (h)
Charlson comorbidity index
Congestive heart failure
Peripheral vascular disease
Chronic lung disease
Connective tissue disease
Mild liver disease
Diabetes without complications
Diabetes with complications
Moderate or severe liver disease
Costs, Length of stay and case weight
Costs for cases with ICU-acquired LRTI and control patients without LRTI
Cases (N = 49)
Controls (N = 49)
Total hospital costs (€)
Reimbursement per patient (€)
23,013 (15,056 - 30,688)
Total ICU costs (€)
11,785 (6,576 - 19,207)
Medical staff (€)
2,506 (1,179 - 4,593)
Nursing staff (€)
4,650 (2,351 - 7,634)
Assistant medical technicians (€)
87 ( 51–178)
Medical products (€)
Daily costs (€)
On the basis of the average € / USD exchange rate for 2010 (1.00 € = 1.39 USD) we converted the median costs into USD. The total hospital costs in USD were 62,607 USD (IQR 42,483 USD - 86,491 USD) for cases and 36,789 USD (IQR 23,490 USD - 36,489 USD) for controls. The attributable costs for an LRTI per patient was 23,651 USD (IQR 6,298 USD - 44.801 USD).
Our study examined costs of ICU-acquired LRTIs in a DRG-based healthcare system in an university hospital. The study population consisted of ICU patients who were admitted to an ICU that participated in the German hospital infection surveillance system. We found that the development of an LRTI resulted in a prolonged hospital stay and higher attributable hospital costs than without a LRTI. Looking at the daily costs, the numbers demonstrate that the average (true) costs per day are also significantly higher in cases compared to controls. These numbers show that a prolonged LOS (even probably the most dominant factor) is not the only reason for additional hospital expenses. The significantly higher case weights for LRTI patients document the increased complexity of these cases even though underlying diseases controlled by SAPS II and Charlson Comorbidity Index were comparable between cases and controls.
Hospital-acquired LRTI is a time dependent event. The longer a patient remains in the hospital, the higher the chances for such an infection. On the other hand, costs in a DRG-system are also modulated by the length of stay. Hence we adjusted our matched cohort by time before onset of LRTI. In this way, we ensured that our data on costs and length of stay would not be overestimated . A study with matched patients tries to compare patients with a similar health condition. There are many scores that can be used to adjust for comorbidities. We used the SAPS II score for matching. We obtained another comorbidity score - the Charlson Comorbidity Index (CCI)  - but we did not use it as a matching criterium. After the matching process, the CCI showed no significant difference and confirmed the successful adjustment for underlying diseases. Among the 90 LRTI cases, we were able to match only 54% due to strict matching criteria. However, we are convinced that we did not overmatch our cohorts since differences in risk factors - like mechanical ventilation time – are still detectable even with statistical significance. Our data therewith affirms the results of other studies that showed mechanical ventilation significantly associated with hospital-acquired LRTI [2, 3, 14]. The ventilation-associated LRTI rate in our cohort was lower than the reference data from Germany (635 ICUs, from 2007 to 2011) which was 4.35 (IQR 2.20-7.30) . The difference could be explained by a strongly preselected patient population that is treated in our university which serves as a highly specialized tertiary care center.
Relevant literature to costs and length of stays (LOS) of patients with lower respiratory tract infection
Adjusted for time – dependency bias
Type of infection
Exposed : unexposed
Attributable hospital LOS
Kollef, 2012 
Matched cohort, retrospectively
2,144 : 2,144
Restrepo, 2010 
Matched cohort, retrospectively
30 : 90
Cocanour, 2005 
Matched case control, prospectively
70 : 70
Warren, 2003 
Cohort study, prospectively
127 : 629
This study, 2012
Matched cohort, prospectively
49 : 49
In our study, an episode of LRTI resulted in an attributable stay of 9 days in the hospital. Even though we did not detect a difference in underlying morbidity LRTI was also associated with a significantly increased mortality. However, our study confirms the results of many other studies that showed that the development of a LRTI leads to a prolonged hospital stay [5, 16–19]. Those studies report an extra hospital stay between 2 and 13 days depending on the study design, the method of calculation and whether the study institutions used the DRG-system. Nosocomial LRTIs are time-dependent events. Hence studies examining LOS need to adjust for this bias . DRG based studies that adjusted for time-dependency reported on median additional hospital stays of 9 to 13 days [6, 7, 17] due to VAP.
There are some limitations to our analysis. Even though we found 90 LRTI patients, we were able to match and analyze only 54% of the cases with non-exposed patients. Therefore the cohort is comparable small. Second, we did not assess the antimicrobial therapy of the analyzed patients. Therefore we could not assess the effect of an adequate and timely antimicrobial therapy on the outcome of our 49 LRTI-patients. Third, we assessed only patients from the Charité University Hospital Berlin. Therefore our results might be representative only for our institution.
The attributable hospital stay for an LRTI was 9 days (p = 0.011) resulting in attributable costs per LRTI of 17,015 € (p < 0.001). Considering the economic impact and the impact on the health-care system, we strongly recommend the introduction of appropriate measurements to prevent the development of hospital acquired LRTI.
- Bonten MJ: Healthcare epidemiology: Ventilator-associated pneumonia: preventing the inevitable. Clin Infect Dis. 2011, 52: 115-121. 10.1093/cid/ciq075.View ArticlePubMedGoogle Scholar
- Joseph NM, Sistla S, Dutta TK, Badhe AS, Parija SC: Ventilator-associated pneumonia: a review. Eur J Intern Med. 2010, 21: 360-368. 10.1016/j.ejim.2010.07.006.View ArticlePubMedGoogle Scholar
- Zilberberg MD, Shorr AF: Economic aspects of preventing health care-associated infections in the intensive care unit. Crit Care Clin. 2012, 28: 89-97. 10.1016/j.ccc.2011.10.005.View ArticlePubMedGoogle Scholar
- Safdar N, Dezfulian C, Collard HR, Saint S: Clinical and economic consequences of ventilator-associated pneumonia: a systematic review. Crit Care Med. 2005, 33: 2184-2193. 10.1097/01.CCM.0000181731.53912.D9.View ArticlePubMedGoogle Scholar
- Cocanour CS, Ostrosky-Zeichner L, Peninger M, Garbade D, Tidemann T, Domonoske BD, Li T, Allen SJ, Luther KM: Cost of a ventilator-associated pneumonia in a shock trauma intensive care unit. Surg Infect (Larchmt). 2005, 6: 65-72. 10.1089/sur.2005.6.65.View ArticleGoogle Scholar
- Kollef MH, Hamilton CW, Ernst FR: Economic impact of ventilator-associated pneumonia in a large matched cohort. Infect Control Hosp Epidemiol. 2012, 33: 250-256. 10.1086/664049.View ArticlePubMedGoogle Scholar
- Restrepo MI, Anzueto A, Arroliga AC, Afessa B, Atkinson MJ, Ho NJ, Schinner R, Bracken RL, Kollef MH: Economic burden of ventilator-associated pneumonia based on total resource utilization. Infect Control Hosp Epidemiol. 2010, 31: 509-515. 10.1086/651669.View ArticlePubMedGoogle Scholar
- Warren DK, Shukla SJ, Olsen MA, Kollef MH, Hollenbeak CS, Cox MJ, Cohen MM, Fraser VJ: Outcome and attributable cost of ventilator-associated pneumonia among intensive care unit patients in a suburban medical center. Crit Care Med. 2003, 31: 1312-1317. 10.1097/01.CCM.0000063087.93157.06.View ArticlePubMedGoogle Scholar
- Gastmeier P, Geffers C, Sohr D, Dettenkofer M, Daschner F, Ruden H: Five years working with the German nosocomial infection surveillance system (Krankenhaus Infektions Surveillance System). Am J Infect Control. 2003, 31: 316-321. 10.1067/mic.2003.66.View ArticlePubMedGoogle Scholar
- Zuschneid I, Rucker G, Schoop R, Beyersmann J, Schumacher M, Geffers C, Ruden H, Gastmeier P: Representativeness of the surveillance data in the intensive care unit component of the German nosocomial infections surveillance system. Infect Control Hosp Epidemiol. 2010, 31: 934-938. 10.1086/655462.View ArticlePubMedGoogle Scholar
- Horan TC, Andrus M, Dudeck MA: CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008, 36: 309-332. 10.1016/j.ajic.2008.03.002.View ArticlePubMedGoogle Scholar
- Beyersmann J, Kneib T, Schumacher M, Gastmeier P: Nosocomial infection, length of stay, and time-dependent bias. Infect Control Hosp Epidemiol. 2009, 30: 273-276. 10.1086/596020.View ArticlePubMedGoogle Scholar
- Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987, 40: 373-383. 10.1016/0021-9681(87)90171-8.View ArticlePubMedGoogle Scholar
- Kollef MH: What is ventilator-associated pneumonia and why is it important?. Respir Care. 2005, 50: 714-721.PubMedGoogle Scholar
- ITS-KISS reference data.http://www.nrz-hygiene.de/surveillance/kiss/its-kiss/.
- Guanche-Garcell H, Requejo-Pino O, Rosenthal VD, Morales-Perez C, Delgado-Gonzalez O, Fernandez-Gonzalez D: Device-associated infection rates in adult intensive care units of Cuban university hospitals: International Nosocomial Infection Control Consortium (INICC) findings. Int J Infect Dis. 2011, 15: e357-e362. 10.1016/j.ijid.2011.02.001.View ArticlePubMedGoogle Scholar
- Kubler A, Duszynska W, Rosenthal VD, Fleischer M, Kaiser T, Szewczyk E, Barteczko-Grajek B: Device-associated infection rates and extra length of stay in an intensive care unit of a university hospital in Wroclaw, Poland: International Nosocomial Infection Control Consortium’s (INICC) findings. J Crit Care. 2012, 27: e105-e110.View ArticleGoogle Scholar
- Mehta A, Rosenthal VD, Mehta Y, Chakravarthy M, Todi SK, Sen N, Sahu S, Gopinath R, Rodrigues C, Kapoor P: Device-associated nosocomial infection rates in intensive care units of seven Indian cities. Findings of the International Nosocomial Infection Control Consortium (INICC). J Hosp Infect. 2007, 67: 168-174. 10.1016/j.jhin.2007.07.008.View ArticlePubMedGoogle Scholar
- Rosenthal VD, Udwadia FE, Munoz HJ, Erben N, Higuera F, Abidi K, Medeiros EA, Fernandez Maldonado E, Kanj SS, Gikas A: Time-dependent analysis of extra length of stay and mortality due to ventilator-associated pneumonia in intensive-care units of ten limited-resources countries: findings of the International Nosocomial Infection Control Consortium (INICC). Epidemiol Infect. 2011, 139: 1757-1763. 10.1017/S0950268811000094.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.