Time to Look for Another Infectious Source? White Blood Cell Trends during Ventilator-Associated Pneumonia

Abstract

Background:

Ventilator-associated pneumonia (VAP) continues to plague patients in intensive care units (ICUs) throughout the world. Persistent leukocytosis despite antibiotic treatment for VAP can have many etiologies including normal inflammatory response, inadequate VAP antimicrobial therapy, and the presence of additional infectious diagnoses.

Hypothesis:

Surgical patients with VAP and a second infectious source have a different white blood cell count (WBC) trend than patients with VAP alone.

Patients and Methods:

Retrospective, single-center study of surgical ICU patients diagnosed with VAP (>10⁴ CFU/mL on semi-quantitative culture) between January 2019 and June 2020. Chart review identified additional infections diagnosed during VAP treatment. White blood cell count values were compared between patients treated for VAP alone (VAP-alone) and those with additional infections (VAP-plus) using a Wilcoxon test. Univariable analysis compared admission type, surgeries, and steroid use between cohorts.

Results:

Eighty-eight VAPs were included for analysis; 61 (69%) were VAP-alone and 27 (31%) VAP-plus. Average age was 47.1 ± 16.7 years, 78% were male, and 93% were trauma admissions. Median hospital day of VAP diagnosis was six (interquartile range [IQR], 4–10). Nearly all patients (99%) were started on initial antibiotic agents to which the VAP organism was sensitive. Daily WBC was higher for VAP-plus compared with VAP-alone on days five, six, and seven of treatment. The maximum WBC was higher for VAP-plus (21.6 k/mcL vs. 16.1 k/mcL; p = 0.02). There were no differences in admission types, number of surgeries, or steroid use between groups.

Conclusions:

Providers should have increased suspicion for additional sources of infection when ICU patients with a VAP continue to have elevated WBC despite appropriate antibiotic therapy.

Pneumonia is a frequent complication for intensive care unit (ICU) patients and is the most common hospital-acquired infection [1,2]. Ventilator-associated pneumonia (VAP) occurs in approximately 10% of patients supported with mechanical ventilation, and despite advances in prevention, this rate has not declined in recent years [3,4]. Ventilator-associated pneumonia contributes substantially to morbidity and mortality rates of ICU patients [5,6] by lengthening the duration of mechanical ventilation, ICU length of stay, and hospital length of stay [2,7,8]. It also adds substantially to hospital costs [9,10].

Pneumonia is diagnosed by a combination of radiographic and clinical evidence [11]. Leukocytosis is an integral part of this diagnosis, as evidenced by its inclusion in the Clinical Pulmonary Infection Score [12] and U.S. Centesr for Disease Control and Prevention's National Healthcare Safety Network ventilator-associated event surveillance [13]. Despite the association of leukocytosis with pneumonia, the duration and amplitude of this laboratory abnormality is unknown [14]. Furthermore, when leukocytosis persists despite appropriate antibiotic therapy, it is unknown if this represents a normal inflammatory response or the presence of additional infectious diagnoses. This study hypothesized that surgical ICU patients with VAP and a second infectious source have a different white blood cell (WBC) count trend than patients with VAP alone.

Patients and Methods

Design and participants

A retrospective study was performed using non-pregnant, adult patients (≥18 years) admitted to Denver Health Medical Center's surgical intensive care unit between January 2019 and June 2020 with a diagnosis of VAP. Patients with additional infection(s) during the treatment of their VAP (VAP-plus) formed the study group while those without additional infection (VAP-alone) became the control group. Patients were excluded from the study if they were incarcerated, died, or were transferred to a different facility before completion of their VAP treatment. If a patient had more than one VAP, each VAP was considered a separate event for analysis. Institutional Review Board approval was obtained for this study.

VAP definition and treatment

Ventilator-associated pneumonia quantitative testing was considered when a patient was mechanically ventilated for at least 48 hours and had a new infiltrate on chest radiograph, increased secretions, or increased oxygenation requirements. This study defined VAP when the quantitative respiratory culture had >10⁴ CFU/mL. Initial antibiotic therapy was started immediately after obtaining the respiratory culture. Our ICU protocol specifies that ceftriaxone should be given empirically when VAP is suspected on hospital days three through five, cefepime for hospital days six through 10, and vancomycin with cefepime for hospital day >10. This regimen is based on previous studies that demonstrate adequacy of coverage [15]. All VAPs were treated with antibiotic therapy for seven days [16].

Data collection and outcome

Data were collected via chart review of the patients' electronic medical records. White blood cell count for each day of VAP treatment was collected. If a patient had multiple WBC values for a single day, the highest value was used.

Additional infections were identified by chart review and were counted if diagnosed during the treatment of the VAP. Bacteremia was included as an additional infection if the pathogen was different from the pneumonia isolate. When patients had a second quantitative respiratory culture sent while receiving treatment for their VAP, this was considered an additional infection if the second culture had >10⁴ CFU/mL and the isolated pathogens required a change in antibiotic agents.

Other data collected from the charts included patients' baseline characteristics and pertinent past medical history that may affect WBC count (active cancer, lymphoma or myeloproliferative disease, history of inflammatory disease or transplant, connective tissue disorder), hospital admission type (trauma, emergency general surgery, elective), surgeries performed, and steroid use. Outcome data gathered included hospital length of stay, ICU length of stay, ventilator days, and all-cause hospital mortality.

Statistical analysis

Data were analyzed using SAS Studio 2021 (Cary, NC). Distribution analysis was performed for all numerical variables using the Kolmogorov-Smirnov test. Non-parametric data were analyzed using the Wilcoxon rank sum test. Categorical data compared between the two groups was completed using χ² test. All non-parametric data are expressed as median and interquartile range, except in graphs where median and 95% confidence intervals are shown. A p value <0.05 was considered statistically significant for all tests. The predictive ability of the maximum WBC count for the outcome of identifying a secondary infectious source was assessed using the area under the receiver operator characteristic curve (AUROC). Youden index was calculated to determine the optimal threshold for the maximum WBC.

Results

Within the study period of January 2019 to June 2020, 87 patients were admitted to the surgical ICU and were diagnosed with a VAP. Four (4.5%) were excluded from the analysis because of incomplete VAP treatment; this left 83 patients who were treated for 88 VAPs. Patients were an average age of 47.1 ± 16.7 years old, 78% male, 84% white, and 30% Hispanic. Two patients had active solid organ cancer (2%) and two had active inflammatory bowel disease (2%). The admission types were 93% trauma (median injury severity score, 26.5), 6% emergency general surgery, and 1% post-operative elective surgery.

The median time to first VAP diagnosis was hospital day six (interquartile range [IQR], 4–10]. For patients treated for a second VAP during the hospital admission, the median time was hospital day 20 (IQR, 13–21). Bacterial isolates of the VAPs in the study were 33% gram positive, 52% gram negative, and 13% both gram positive and gram negative. Nearly all (87/88; 99%) of the empiric antibiotic agents begun for VAP covered the isolated pathogens. Ventilator-associated pneumonia alone had more gram-positive bacterial isolates than the VAP-plus cohort (41% vs. 15%; p = 0.03).

There were 61 VAPs in the VAP-alone cohort (69%) and 27 in the VAP-plus cohort (31%). The characteristics of the two cohorts are shown in Table 1. The additional infections identified in the VAP-plus group were eight (27%) intra-abdominal infections; six (20%) Clostridium difficile infections; six (20%) with an untreated respiratory pathogen; four (13%) skin, soft tissue, or superficial wound infections; two (7%) urinary tract infections; two (7%) bacteremia or fungemia; and two (7%) pleural space infections. Three patients in the VAP-plus cohort had more than one additional infection found during their VAP treatment. Diagnosis of the additional infection occurred on average day five ± two of VAP treatment.

Table 1.

Comparison of VAP-Alone and VAP-Plus Cohorts with Data Displayed as Median (IQR) or Count (Percentage of Total)

	VAP-alone (n = 61)	VAP-plus (n = 27)	p
Hospital day of first VAP diagnosis	6 (3,10)	6 (4,10)	0.87
ICU day of VAP diagnosis	6 (3,10)	6 (4,10)	0.87
VAP pathogens
Gram positive only	25 (41%)	4 (15%)	0.03
Gram negative only	28 (46%)	18 (67%)	0.10
Gram positive and negative	8 (13%)	3 (11%)	1.00
Fungal	5 (8%)	2 (7%)	1.0
Median WBC during VAP treatment (k/mcL)
Day 1	12.2 (9.3,16.3)	15.5 (8.1,18.7)	0.51
Day 2	11.4 (9.4,17.6)	13.3 (8.5,20.1)	0.83
Day 3	12.8 (9.0,16.1)	13.2 (8.5,18.2)	0.58
Day 4	11.8 (8.6,14.5)	13.3 (9.1,19.4)	0.16
Day 5	12.1 (9.0,14.7)	14.3 (10.8,22.7)	0.007
Day 6	12.0 (10.0,15.9)	16.3 (13.0,23.9)	<0.001
Day 7	12.7 (9.9,5.1)	16.3 (12.9,22.6)	0.002
Maximum during day 1–7	16.1 (12.7,21.1)	21.6 (13.3,26.7)	0.02
Corticosteroid use (2 d before VAP through end of VAP treatment)	8 (16%)	3 (11%)	1.00
Surgery performed (2 d before VAP through end of VAP treatment)	32 (51%)	20 (74%)	0.06
Intra-cranial	4 (7%)	2 (7%)	1.0
Spinal	5 (8%)	1 (4%)	0.66
Tracheostomy	14 (23%)	11 (41%)	0.12
Thoracic	1 (2%)	3 (11%)	0.08
Abdominal/pelvic	0 (0%)	4 (15%)	0.008
Orthopedic	11 (18%)	3 (11%)	0.54
Skin and soft tissue	4 (7%)	1 (4%)	1.00
Other type	11 (18%)	3 (11%)	0.54

VAP = ventilator-associated pneumonia; ICU = intensive care unit; IQR = interquartile range.

The median WBC count for each day of VAP treatment (day one through day seven) for the two cohorts are shown in Figure 1 and Table 1. The VAP-plus cohort had higher median WBC counts on days five through seven of VAP treatment (p = 0.007, < 0.001, 0.002). The maximum WBC count during VAP treatment was higher in the VAP-plus (21.6 k/mcL) than the VAP-alone cohort (16.1 k/mcL; p = 0.02). Concomitant corticosteroid treatment and surgery during VAP treatment was the same for both groups, except that the VAP-plus cohort had statistically significant more abdominal surgery than the VAP-alone group (15% vs. 0%; p = 0.008). No comparison of the neutrophil or immature neutrophils percentages between cohorts was performed because only 7.5% of the CBCs were sent with differential counts. The AUROC of the maximum WBC for identifying a second infection in patients being treated for VAP was 0.66, with an optimal cutoff of 20.1 k/mcL, and a Youden index of 0.35 (sensitivity of 63% and specificity of 72%; p = 0.318).

Fig. 1.

Median white blood cell count (WBC) of ventilator-associated pneumonia (VAP)-alone and VAP-plus cohorts by day of VAP antibiotic treatment. Shaded areas are the 95% confidence intervals.

The VAP-plus and VAP-alone cohorts had different hospital outcomes. The VAP-plus had a longer median hospital length of stay (33 vs. 24 days; p = 0.045) and longer median ICU length of stay (22 vs. 16 days; p = 0.013). Ventilator-free days were similar between groups (VAP-plus 12 days vs. VAP-alone 14 days; p = 0.332). All-cause hospital mortality was 20% for the VAP-plus group and 12.5% for the VAP-alone, p = 0.346.

Discussion

We identified that WBC counts for ICU patients with VAP and second infectious source are different than patients with only VAP. The maximum WBC during the VAP treatment was higher as was the daily WBC count on days five through seven of treatment. This information may be useful to ICU clinicians who are trying to determine when additional infectious evaluation is needed for patients already diagnosed with a VAP. Testing for infectious sources add hospital costs [17] and are often low yield [18 –20]; knowing that a persistent leukocytosis may be a sign of an untreated infectious source may help inform clinical decision-making.

Leukocytosis is a relatively common finding with a wide differential, found in both infectious and non-infectious inflammatory states. It can be observed secondary to bacterial infections, but also occurs after trauma, surgical intervention, and inflammatory processes, which are also common in the surgical ICU. Any source of stress can cause a catecholamine-induced demargination of WBCs, as well as increase the release from the bone marrow storage pool [21].

Previous studies have shown that leukocytosis alone is a poor predictor of infection [22]. In a systemic review of adult immunocompentent patients, isolated leukocytosis had a likelihood ratio of <1.7 for predicting bacteremia [23]. In acute appendicitis, Yu et al. [24] found the predictive value of leukocytosis had only a sensitivity and specificity of 62% and 75%. It has also been studied its ability to diagnosis of VAP [14] and urinary tract infections [18] and found not to have predictive capability.

The degree and duration of leukocytosis has also not previously been shown to be indicative of an infectious process. Potasman and Grupper [25] studied 173 adult patients without hematologic malignancies with WBC counts >30,000 k/mcL and found that infection was the common cause in 48%, followed by ischemia (28%), inflammation (7%), and obstetric diagnoses (7%). Higher WBC counts were associated with positive blood cultures or a positive Clostridium difficile toxin [25]. Haburchak and Alcheiki [26] looked at infectious disease consults for unexplained and persistent leukocytosis and found most patients had extensive tissue damage rather than active infection driving the leukocytosis. These studies contrast the results of this study, where there was an increased leukocytosis with the VAP-plus cohort presumably due to the additional infections had by these patients. The reason for these differences is likely due to the way patients were selected: the previous studies selected based on leukocytosis whereas this study selected based on infection. There was also a wide range in the WBC counts of the VAP-plus cohort, but given the low study size, it was not possible to make more meaningful subanalysis of characteristics that drove the increased leukocytosis (e.g., type of infection, type of traumatic injury).

It is possible that persistent leukocytosis observed in this study could be part of a persistent inflammation, immunosuppression and catabolism syndrome (PICS) [27]. This syndrome is observed in patients with prolonged surgical ICU stays and characterized by persistent immune dysfunction resulting in multiple-organ failure, increased catabolism, and recurrent infections. Although PICS is relatively uncommon [28], the study cohort did include patients with multiple infections and had longer than average ICU stays secondary to their VAP. Unfortunately, data regarding patients' catabolic state was not consistently recorded in the medical record and thus we were not able to determine which patients in the study could be classified as having PICS.

There are multiple limitations to this study. First, the data was from a single ICU that cares for a high percentage of trauma patients, which may not be representative of other ICUs. Others have shown that trauma populations may behave different from other ICU populations in terms of VAP given their altered immune function secondary to traumatic injury [29,30]. The retrospective design created bias. Specifically, if a patient had a persistently elevated WBC count, it is possible that this prompted the team to pursue additional infectious workup compared to those patients whose WBC normalized after the start of antibiotics. Although an infectious workup does not always result in a new infectious diagnosis, it intuitively will find more infections compared to patients without one. During the study period, there was no protocol for when infectious workup was performed; commonly it was based on fever (38.5°C), increasing WBC count, or worsening clinical status. Also, this study only looked at infections detected during the seven-day VAP treatment; if the presentation of the additional infection was identified after the VAP treatment concluded, it would have been excluded from this analysis. Further studies involving multiple centers would be helpful to confirm the findings of this study.

Conclusions

Surgical ICU patients with VAP and an additional infectious source had higher WBC counts on antibiotic treatment days five through seven and higher maximum WBC counts than patients with VAP alone. Providers should have increased suspicion for additional sources of infection when patients' WBC counts remains elevated despite appropriate antibiotic therapy.

Footnotes

Authors' Contributions

Conceptualization: Werner. Conceptualization (supporting): Burlew. Data curation: Werner. Formal analysis (lead): Cralley. Formal analysis: Pieracci. Writing (lead): Werner. Review and editing of final draft (lead): Werner. Review and editing (equal): Cralley, Lawless, Platnick, Cohen, Coleman, Hoehn, Campion, Pieracci, Burlew.

Funding Information

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Author Disclosure Statement

All authors declare no competing or personal financial interests, or any other competing interests.

References

Magill

, O'Leary

, Janelle

, et al. Changes in Prevalence of health care-associated infections in U.S. hospitals. N Engl J Med, 2018; 379:1732–1744.

Papazian

, Klompas

, Luyt

. Ventilator-associated pneumonia in adults: a narrative review. Intensive Care Med. 2020; 46:888–906.

Wang

, Eldridge

, Metersky

, et al. National trends in patient safety for four common conditions, 2005–2011. N Engl J Med, 2014; 370:341–351.

Metersky

, Wang

, Klompas

, et al. Trend in Ventilator-associated pneumonia rates between 2005 and 2013. JAMA, 2016; 316:2427–2429.

Steen

, Vansteelandt

, De Bus

, et al. Attributable mortality of ventilator-associated pneumonia. Replicating findings, revisiting methods. Ann Am Thorac Soc, 2021; 18:830–837.

Bekaert

, Timsit

, Vansteelandt

, et al. Attributable mortality of ventilator-associated pneumonia: A reappraisal using causal analysis. Am J Respir Crit Care Med, 2011; 184:1133–1139.

Muscedere

, Day

, Heyland

. Mortality, attributable mortality, and clinical events as end points for clinical trials of ventilator-associated pneumonia and hospital-acquired pneumonia. Clin Infect Dis, 2010; 51(Suppl 1):S120–125.

Luo

, Xing

, Wang

. The effect of ventilator-associated pneumonia on the prognosis of intensive care unit patients within 90 days and 180 days. BMC Infect Dis, 2021; 21:684.

Zimlichman

, Henderson

, Tamir

, et al. Health care-associated infections: A meta-analysis of costs and financial impact on the US health care system. JAMA Intern Med, 2013; 173:2039–2046.

10.

Kollef

, Hamilton

, Ernst

. Economic impact of ventilator-associated pneumonia in a large matched cohort. Infect Control Hosp Epidemiol, 2012; 33:250–256.

11.

Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare-associated pneumonia. Am J Respir Crit Care Med 2005;171:388–416.

12.

Pugin

, Auckenthaler

, Mili

, et al. Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic “blind” bronchoalveolar lavage fluid. Am Rev Respir Dis, 1991; 143:1121–1129.

13.

Magill

, Klompas

, Balk

, et al. Developing a new, national approach to surveillance for ventilator-associated events. Crit Care Med, 2013; 41:2467–2475.

14.

Klompas

. Does this patient have ventilator-associated pneumonia?. JAMA, 2007; 297:1583–1593.

15.

Lollar

, Rodil

, Herbert

, et al. Empiric methicillin resistant Staphylococcus aureus coverage in the early ventilator associated pneumonia window: If and when. Surg Infect, 2016; 17:187–190.

16.

Kalil

, Metersky

, Klompas

, et al. Management of adults with hospital-acquired and ventilator-associated pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis, 2016; 63:e61–e111.

17.

Russo

, Landon

, Burkard

, et al. Fever evaluation cost and time to treatment in the general surgery patient can be reduced by using a fever practice guideline. Mil Med, 2014; 179:1166–1170.

18.

Golob

Jr. , Claridge

, Sando

, et al. Fever and leukocytosis in critically ill trauma patients: It's not the urine. Surg Infect, 2008; 9:49–56.

19.

Claridge

, Golob

Jr , Fadlalla

, et al. Fever and leukocytosis in critically ill trauma patients: It is not the blood. Am Surg, 2009; 75:405–410.

20.

Laupland

, Shahpori

, Kirkpatrick

, et al. Occurrence and outcome of fever in critically ill adults. Crit Care Med, 2008; 36:1531–1535.

21.

Riley

, Rupert

. Evaluation of patients with leukocytosis. Am Fam Physician, 2015; 92:1004–1011.

22.

Van den Bruel

, Thompson

, Haj-Hassan

, et al. Diagnostic value of laboratory tests in identifying serious infections in febrile children: Systematic review. BMJ, 2011; 342:d3082.

23.

Coburn

, Morris

, Tomlinson

, Detsky

. Does this adult patient with suspected bacteremia require blood cultures?. JAMA, 2012; 308:502–511.

24.

, Juan

, Wu

, et al. Systematic review and meta-analysis of the diagnostic accuracy of procalcitonin, C-reactive protein and white blood cell count for suspected acute appendicitis. Br J Surg, 2013; 100:322–329.

25.

Potasman

, Grupper

. Leukemoid reaction: Spectrum and prognosis of 173 adult patients. Clin Infect Dis. 2013; 57:e177–181.

26.

Haburchak

, Alchreiki

. Inpatients with 'unexplained' leukocytosis. Am J Med, 2020; 133:508–514.

27.

Gentile

, Cuenca

, Efron

, et al. Persistent inflammation and immunosuppression: A common syndrome and new horizon for surgical intensive care. J Trauma Acute Care Surg, 2012; 72:1491–1501.

28.

Hesselink

, Hoepelman

, Spijkerman

, et al. Persistent inflammation, immunosuppression and catabolism syndrome (PICS) after polytrauma: A rare syndrome with major consequences. J Clin Med. 2020; 9:191.

29.

Koulenti

, Tsigou

, Rello

. Nosocomial pneumonia in 27 ICUs in Europe: Perspectives from the EU-VAP/CAP study. Eur J Clin Microbiol Infect Dis, 2017; 36:1999–2006.

30.

Cook

, Norwood

, Berne

. Ventilator-associated pneumonia is more common and of less consequence in trauma patients compared with other critically ill patients. J Trauma, 2010; 69:1083–1091.