Surveillance for Ventilator-Associated Pneumonia: Can We Apply Centers for Disease Control and Prevention–National Healthcare Safety Network 2013 Definitions for All Settings?

To the Editor—In the present journal, we read the article by Greene et al ¹ about the influence of Centers for Disease Control and Prevention–National Healthcare Safety Network (CDC-NHSN) 2015 definitions for catheter-associated urinary tract Infection surveillance with great interest. Here, we share our experience of using CDC-NHSN 2008 and 2013 criteria for ventilator-associated pneumonia (VAP) surveillance at different intensive care unit (ICU) settings.

VAP is one of the most important problems in ICUs. Adequate surveillance of VAP is critical in order to introduce effective control measures early. The surveillance criteria for VAP was released by CDC-NHSN in 2008 and 2013. ^{2 , 3} The 2013 criteria was based on the worsening pulmonary functions. The worsening oxygenation—as increase in positive end-expiratory pressure and fraction of inspired oxygen—was termed ventilator-associated condition. When an abnormal temperature or white blood cell count and new antibiotics are added to a ventilator-associated condition, this condition is described as infection-related ventilator-associated complication. Diagnosis of possible VAP requires detection of specific microbioogic etiology in addition to an infection-related ventilator-associated complication. From January 1, 2013, through March 30, 2015, we adapted CDC-NHSN 2013 definitions for VAP surveillance; however, we observed a huge difference between the number of patients with a clinical diagnosis of VAP versus the VAP rate detected by CDC-NHSN 2013 criteria, pariticularly in the surgical ICUs. Then, we decided to use CDC-NHSN 2008 criteria for VAP surveillance to understand the role of definitions in the rate of the rapidly changed VAP rate.

Our hospital is a 700-bed tertiary center. A patient-based infection control program has been set up more than 20 years. An infection control nurse daily visits all patients hospitalized in the ICU to detect ICU-acquired infections. All surveillance data are periodically discussed with the infection control doctor, who is an infectious diseases physician. There are 3 surgical ICUs with 42 beds (general surgery with 9 beds, cardiothoracic surgery with 22 beds, and neurosurgery with 11 beds), and a medical ICU (MICU) with 9 beds in our adult hospital. In surgical ICUs, there are no intensivists and all patients are followed by individual surgical teams who operate on the patients. Owing to lack of adequate number of staff, ventilator parameters are not recorded properly and even sometimes cannot be managed according to the actual clinical condition of the patients. Diagnostic tests such as complete blood count or cultures from respiratory tract and blood are delayed because of shortage of well-trained staff in surgical ICUs. Intensivists are available for 24 hours in the MICU in contrast to the surgical ICUs. Mechanical ventilation support is managed according to the actual respiratory condition of the patient and all changes are recorded properly. Diagnostic tests are performed in a timely manner when there is a suspicion of ICU-acquired infection. Although there is a weaning algorithm in the MICU, it does not exist in surgical ICUs.

A total of 3,516 ventilation-days were followed from January 1, 2013, through December 31, 2013, in 4 ICUs. The VAP rates detected by CDC-NHSN 2008 criteria were as follows, per 1,000 ventilation-days: 9.6 in MICU, 9.9 in general surgery ICU, 19.2 in cardiothoracic ICU, and 19.1 in neurosurgical ICU.

During the second period, CDC-NHSN 2013 criteria was adapted for VAP surveillance and 4,298 ventilation-days were followed from January 1, 2014, through March 30, 2015. Diagnosis of possible VAP was confirmed by Ventilator-Associated Event Calculator, version 2.1 (CDC), during this period. The VAP rate was suddenly decreased in all surgical ICUs (from 9.9 to 1.6 in general surgery ICU, from 19.2 to 1.7 in cardiothoracic ICU, and from 19.1 to 4.3 in neurosurgical ICU) and a slight decrease was observed in MICU (from 9.6 to 7.5) (Table 1).

TABLE 1 Rates of Ventilator-Associated Pneumonia (VAP) According to Type of ICU

NOTE. VAP rate was calculated for 1,000 ventliation-days. ICU; ıntensive care unit.

From April 1, 2015, through December 30, 2015, a total of 2,747 ventilation-days were followed. During this period all surgical ICUs had a sharp increase for VAP rates in surgical ICUs (from 1.6 to 9.8 in general surgery ICU, from 1.7 to 22.5 in cardiothoracic ICU, and from 4.3 to 16.8 in neurosurgical ICU) whereas the VAP rate in MICU was stable after reapplying the CDC-NHSN 2008 criteria for VAP surveillance. Ventilator utilization rates were similiar for all periods when intra-unit comparison was performed (Table 1).

Our study was not a head-to-head comparison of CDC-NHSN 2008 and 2013 criteria. However, a recent study reported that VAP rate showed a significant difference when different criteria was used. This multicenter study showed that 2 of the 4 ICUs achieved zero VAP rates using the CDC-NHSN 2013 algorithm. ⁴

Subjective interpretation of radiologic findings of pneumonia is the major disadvantage of CDC-NHSN 2008 criteria of VAP surveillance. The primary aim of the CDC-NHSN 2013 criteria was to avoid the interpretation variability. However, it may not be so easy to follow the CDC-NSHN 2013 algorithm for VAP surveillance when the infrastructrure of the ICU has certain limitations. Our findings show the importance of addressing the limited usefulness of CDC-NHSN 2013 in settings where human resources are sparse. Potential failure to capture VAP can be an important problem for effective control of ICU-acquired infections.