Hospital-acquired pneumonia (HAP) is now the most common type of hospital-acquired infection in the United States, accounting for 26% of all such infections, according to the most recent point prevalence survey conducted by the Centers for Disease Control and Prevention (CDC).1 It is also the most common hospital-acquired infection in Europe.2, 3 Of the HAP cases in the CDC survey, fully 65% were found to be nonventilator hospital-acquired pneumonia (NV-HAP).
NV-HAP presents a serious and largely preventable threat to patient safety. Compared with all other hospital-acquired infections except ventilator-associated pneumonia, NV-HAP is associated with increased morbidity and mortality, longer hospital stays, increased ICU utilization, and markedly higher costs.4-7 One study found that, after controlling for multiple comorbidities and other factors, patients with NV-HAP were at much higher risk for death during hospitalization than those without NV-HAP (15.5% versus 1.6%).7
Pathogenesis. Pneumonia occurs when microbes move from proximal sites such as the oral cavity into the lungs and incite an inflammatory response.8-10 Researchers have found an important relationship between the oral microbiota and HAP.10-12 For example, bacteria found in the bronchi of patients with HAP have been matched with specific flora found in the oral cavity.8, 13, 14 Moreover, within 48 hours of hospitalization, changes occur in the oral microbiota that are associated with more virulent pneumonia-causing organisms.15, 16 Respiratory pathogens such as Staphylococcus aureus, Klebsiella pneumoniae, Enterobacter cloacae, and Pseudomonas aeruginosa colonize the dental plaque and can be aspirated, introducing these organisms into the lungs, even in healthy adults.17-19 Studies indicate that patients are up to six times more likely to develop HAP if the mouth is persistently colonized by such pathogens.15, 20 Several studies have shown that improved cleaning of the oral biofilm provides primary source control of HAP pathogens and reduces HAP incidence.5, 14, 21
Given the relatively high risk of NV-HAP, it's imperative that we gain a better understanding of effective means of prevention. An emerging body of literature addresses the role of oral care in preventing NV-HAP.4, 21-26 An associated reduction in antibiotic use has also been noted.27 Because oral microbiota is arguably the most modifiable risk factor for NV-HAP,21, 22 primary source control was the underlying strategy used in this study.
Study aim. The primary aim of this study was to determine the effectiveness of a universal, standardized oral care protocol in preventing NV-HAP in the acute care setting, specifically, four units at an 800-bed tertiary medical center. We hypothesized that there would be a significant reduction in NV-HAP incidence on the intervention units compared with the control units. Frequency of oral care in compliance with the new oral care protocol was used to assess fidelity to the intervention. The primary outcome measure was NV-HAP incidence per 1,000 patient-days. Our secondary aim was to learn more about NV-HAP outcomes and add this to the emerging literature.4, 6, 28 To that end, we also collected data on NV-HAP–associated events, including the development of postadmission sepsis, unplanned ICU transfer, 30-day readmission, and mortality.
Study design and setting. The study was designed as a single-site cluster randomized trial. It involved a total of four clinical units: one medical and one surgical unit were randomized to an enhanced oral care protocol; these were matched to one medical and one surgical unit randomized to usual care. The study units were selected based on similar patient admission diagnoses and incidence of NV-HAP as evaluated over the preceding 12 months. Unit randomization was done by simple, blind drawing from a container. The medical intervention and control units had 49 and 38 beds, respectively. The surgical intervention unit had 26 beds for postoperative cardiac surgery patients; the surgical control unit had 32 beds for postoperative vascular surgery patients. Nurse-to-patient ratios were 1:4 on the medical units and 1:3 on the surgical units. Nursing assistant-to-patient ratios were 1:9 on the medical units and 1:7 on the surgical units. Other nursing care actions that might influence NV-HAP rates, such as head of bed elevation and early mobilization, were not changed and were similar between the control and intervention units.
Sample. To determine the necessary sample size, a power analysis was conducted via G∗Power 22.214.171.124 freeware.29 Using data from previous research on NV-HAP incidence and preimplementation data, we performed the analysis with a target set at 50% reduction in the NV-HAP incidence rate from 1.8 per 1,000 patient-days (control) to 0.9 per 1,000 patient-days (intervention). Results indicated that 2,580 patients per group—a total of 10,320 patients—were needed to power the study, per the typical settings of α = 0.05 and power = 0.8 (α represents the probability of finding significance where there is none; power represents true significance).29
Approval for the study was obtained from the organization's institutional review board prior to data collection. Signed consent was waived because the intervention presented minimal risk to participants.
The intervention consisted of the implementation of a standardized oral care protocol (see Figure 1) and targeted education. The protocol, which was first developed for an earlier study,30 was adapted by two of us (DB and KKG) for this study. It was approved by the American Dental Association (ADA) Board of Trustees in 2017. Before implementation, all nurses and nursing assistants on the intervention units received education on the protocol. In individual and group sessions, the investigators covered topics that included the importance of oral care, methods of delivery, and assessment of patient needs. Demonstrations of new oral care techniques and products used in the study were given, and regularly scheduled coaching as well as reinforcement when needed were offered throughout the study period. Patients and family members on the intervention units were educated through flyers, signage, teaching that emphasized the importance of oral care, and demonstrations of oral care techniques and products.
The intervention was designed to align with the ADA-approved protocol's recommendations. It made use of a kit that included a high-quality soft-bristle toothbrush, plaque-removing toothpaste, a nonalcohol-based mouthwash, and lip and mouth moisturizers applied as needed. Target frequency of oral care was set at four times a day.30 The frequency and timing were similar whether the patient was on a regular meal schedule or not. Suction toothbrush kits were used with patients who were at risk for aspiration or unable to perform their own oral care. Patients with dentures were given special kits that included a high-quality soft-bristle toothbrush, denture adhesive cream, and a denture cup and sanitizing tablets for nightly cleaning. Every kit came with standard directions for use and additional patient education materials developed for the study.
The kits used in the study were provided by Medline Industries (Northfield, IL). They were chosen because they were the only kits available that contained all the products in the protocol. (Organizations can also follow the protocol's recommendations by assembling kits using individual products from a variety of manufacturers.) Unit-level team members (nurses and nursing assistants) were instructed to perform daily oral assessments to ensure that patients had the correct supplies based on their status, which could change. They were asked to document both these assessments and oral care in the electronic health record (EHR).
On the control units, all unit-level team members received refresher training on the hospital's usual oral care protocol and required documentation. No special oral care supplies were provided, but team members were reminded to document all oral care performed. In keeping with usual care, patients were not reminded to brush their teeth or educated on the importance of doing so. Assistance with oral care was provided as needed. No attempt was made to inform team members on the control units of the procedures and materials used on the intervention units.
Determining NV-HAP incidence. All 8,713 patients discharged from the four study units during the study period (a number that includes all those who died) were screened for inclusion. Those who weren't at least 18 years of age were excluded. Cases of pneumonia were then identified using the International Classification of Diseases, Tenth Revision (ICD-10) codes, as documented in the hospital's electronic data warehouse. Patients without such codes were excluded, as were those who had been hospitalized less than 48 hours, received a pneumonia diagnosis within 48 hours of admission, been extubated within 48 hours of pneumonia diagnosis, or been present on a study unit less than 48 hours before such diagnosis. From the remaining cases, NV-HAP was confirmed using the CDC's criteria for pneumonia (which involve radiologic determination, cultures when available, and symptom assessment).31
Data collection. After staff education on the new oral care protocol was completed, there was a 30-day run-in period (September 1 through September 30, 2018) to monitor compliance and provide additional training as needed. Data collection took place over a 12-month period, from October 1, 2018, through September 30, 2019.
For all patients in the study, frequency of oral care was recorded by nurses and nursing assistants in the EHR. Because the EHR did not permit documentation at the level of detail needed to track the new oral care protocol, randomized audits were conducted weekly via direct patient interviews on all study units. These took place in real time with an average of 63% of patients on a unit, depending on census. An investigator (AM) asked patients and family members about their experience in the last 24 hours: the number of times they'd completed oral care, the level of assistance needed, and whether they'd used their own oral care products or those provided by the hospital.
Data analysis. All data were entered into an Excel spreadsheet and were audited for accuracy by two clinical research coordinators (AM and another coordinator at the hospital) and by the research team. Data were then imported into IBM SPSS, version 26, for analyses. The NV-HAP incidence rate per 1,000 patient-days on the intervention units was compared with that on the control units. A Χ2 analysis was used to determine the relationship between units, and the Wald test for logistic regression was calculated to obtain the odds ratio and confidence intervals between groups. Means comparisons of continuous variables (age, length of stay) across the control and intervention groups were done using the independent samples t test. The Mann-Whitney U test was used to test for group differences of categorical variables (gender, race). The Cramér V effect size was calculated to assess the strength of association between categorical variables. Monthly tracking of NV-HAP cases, compliance with oral care practice standards, and pneumonia readmission rates on all study units were summarized using descriptive statistics to compare the control and interventional units. For all analyses, statistical significance was set at P ≤ 0.05.
Sample. Of the 8,713 patients initially screened, four were excluded because they were under 18 years of age, leaving a total of 8,709 patients. Distribution across the study units was as follows: medical control, 2,075 patients; surgical control, 2,088 patients; medical intervention, 2,709 patients; surgical intervention, 1,837 patients. All 8,709 patients were evaluated for the presence of NV-HAP using the criteria described above. At the end of data collection, nearly three times as many NV-HAP cases had been identified in the control group (29 patients) as in the intervention group (10 patients). See Figure 2 for a flow diagram describing our process for obtaining this final patient sample.
Regarding age, there were small but significant differences between both the medical and surgical control versus the intervention groups. Patients in the medical and surgical control groups had a mean age of 60.9 and 62.6 years, respectively. Those in the medical and surgical intervention groups had a mean age of 62 and 64.9 years, respectively. Regarding length of stay, there was a significant difference between the surgical control group (mean, 5.3 days) and the surgical intervention group (mean, 7.4 days). Regarding race, there were no significant differences between the medical control and intervention units. On the surgical units, there were significantly more white patients on the intervention unit (64.5%) than on the control unit (50.1%) and significantly more Asian patients on the control unit (27.1%) than on the intervention unit (12.2%). Demographic data for the full sample are provided in Table 1.
Table 1. -
Patient Demographics (N = 8,709) by Unit
|Control (n = 2,075)
||Intervention (n = 2,709)
||Control (n = 2,088)
||Intervention (n = 1,837)
|Age, mean (SD), years
|| 60.9 (17.6)
|LOS, mean (SD), days
|| 5.5 (6.7)
|Gender, No. (%)
|Race, No. (%)a
| Black or AfricanAmerican
LOS = length of stay.
aPercentages are based on total number per unit. Some data were missing; thus, not all columns sum to 100%.
Deidentified demographic and general outcomes data were retrieved for the 39 identified NV-HAP patients. Demographic data included age, gender, length of stay, and hospital day of NV-HAP onset. Outcomes data included sepsis diagnosis postadmission, unplanned ICU transfer, 30-day readmission, and mortality. For details, see Table 2.
Table 2. -
Demographic and Outcomes Data for NV-HAP Patients (n = 39)
|Control (n =16)
||Intervention (n = 3)
||Control (n = 13)
||Intervention (n = 7)
| Mean (SD)
|Gender, No. (%)
|| 9 (56.2)
|| 2 (66.7)
|| 6 (46.2)
|| 2 (28.6)
|| 7 (43.8)
|| 1 (33.3)
|| 7 (53.8)
|| 5 (71.4)
| Mean (SD)
|Hospital day of NV-HAP onset
| Mean (SD)
|| 8 (6)
|| 5 (1)
|| 6 (2)
|| 11 (5)
Outcomes for All NV-HAP Cases, No. (%)
Sepsis Diagnosis Postadmission
Unplanned ICU Transfer
LOS = length of stay; NV-HAP = nonventilator hospital-acquired pneumonia.
Primary outcome. To determine the effectiveness of a universal, standardized oral care protocol in preventing the development of NV-HAP, we looked at the incidence rate of NV-HAP per 1,000 patient-days. Logistic regression revealed a significant difference between the medical control and intervention groups. Patients on the medical control unit were 7.1 times more likely to develop NV-HAP than those on the medical intervention unit. The Cramér V effect size was 0.52, indicative of a moderate association between being on a control unit and developing NV-HAP. But the difference between the surgical control and intervention groups did not reach significance. For all units, the NV-HAP incidence rate per 1,000 patient-days and the percentage by which this decreased are shown in Table 3.
Table 3. -
NV-HAP Rates and χ2
Results by Group
||Incidence Rate per 1,000 Patient-Days
|Medical Control, No. (%)
|Medical Intervention, No. (%)
||−85 (% difference)
|Surgical Control, No. (%)
|Surgical Intervention, No. (%)
|| 7 (0.4)b
||−56 (% difference)
CI = confidence interval; NV-HAP = nonventilator hospital-acquired pneumonia; OR = odds ratio.
aOR for medical control vs. medical intervention units (OR: 7.1; 95% CI, 2.01-24.1, P = 0.002).
bOR for surgical control vs. surgical intervention units (OR: 1.6; 95% CI, 0.65-4.1, P = 0.29).
The initial power analysis indicated a requisite sample size of 2,580 patients per group, for a total of 10,320 patients. While none of the four groups reached a sample size of 2,580, post hoc power analyses of the medical control versus the medical intervention units revealed a moderate Cramér V effect size of 0.54 and an actual power of 0.99. However, such analyses of the surgical control versus the surgical intervention units revealed a much smaller Cramér V effect size of 0.02 and a lower power of 0.61.
While our goal of achieving an oral care frequency of four times daily wasn't met on any of the study units, the frequency was significantly higher on both the intervention units compared with the control units. On the medical units, the intervention unit increased the daily oral care frequency by 135% compared with the control unit. On the surgical units, the intervention unit increased the frequency by 70% compared with the control unit. For more details on oral care frequency and types of oral care provided, see Table 4.
Table 4. -
Oral Care Frequency and Type Provided
|Frequency of Daily Oral Care per Month
|Control, range (mean)
||0.86 -1.03 (0.95)
||0.9 -1.4 (1.18)
|Intervention, range (mean)
||1.28 - 2.9 (2.25)
||1.19 - 2.3 (2.02)
Type of Oral Care Provided, % of Patients
Required some assistance
Required total assistance and use of suction toothbrush
This study tested the impact of enhanced oral care on NV-HAP prevention in patients on medical and surgical units. Over the course of the study, the daily oral care frequency improved significantly on the intervention units, reaching up to 2.9 times per day on the medical intervention unit. Although this didn't meet the target level, improved oral care was associated with NV-HAP reductions of 85% on the medical intervention unit and 56% on the surgical intervention unit. This finding supports previous research that showed a positive relationship between increased frequency of daily oral care and NV-HAP incidence reduction.23
Studies of implementation science can involve a variety of contextual factors that aren't specifically related to the intervention itself, including nursing culture and leadership, workload, administrative responsiveness and support, and level of resistance to change.32-34 We found this to be so in our study. The medical intervention unit embraced the oral care intervention and provided more frequent oral care (135% more than control) compared with the surgical intervention unit (70% more than control)—even though the medical unit had less staffing. Also, more patients on the medical than the surgical intervention unit were identified as partially or entirely dependent on assistance for oral care (59% versus 44%). There was a general expectation that patients could perform oral self-care in the first days after cardiovascular surgery, and this may have accounted for the lower frequency of oral care on the surgical intervention unit compared with the medical intervention unit. The surgical intervention unit also had a significantly higher mean length of stay and mean patient age compared with the surgical control unit; both variables are known risk factors for hospital-acquired infection.
Post hoc power analyses showed that the surgical units were underpowered to answer the research question, and the effect size was very small. Even with 12 months of data collection, we were unable to achieve the full sample of patients.
Despite the use of multiple means of education, coaching at the point of care, readily available supplies, and ongoing reports during weekly huddles, oral care frequency remained lower than our target level. This is consistent with the findings of other researchers exploring the impact of oral care on NV-HAP rates and the challenges of improving such care.21, 23, 35, 36
As oral care is solely a nursing intervention—and one critical to patient safety—nurses are well positioned to have a strong impact in this area. Nursing assistants are also vital to oral care improvement, as they often provide much of this basic care. In order to realize the importance of the nursing role in NV-HAP prevention, we must change the mindset of nurses and other health care providers from seeing oral care as a comfort measure to recognizing oral care as a therapeutic intervention and oral care products as therapeutic devices. From an organizational perspective, there is evidence that the return on investment in higher-quality oral care products for use in NV-HAP prevention efforts can be substantial.21 But this requires a sizable initial investment in such products, which will present a roadblock for many hospitals. In short, changing the clinical mindset will take time, sustained effort, ongoing involvement of nurses and nursing assistants, interdisciplinary collaboration, and buy-in from nursing and hospital leadership.
Although we collected demographic data on age, gender, and race, the small number of NV-HAP cases did not permit analyses to assess these data in relation to findings. The influence of age, gender, and race on NV-HAP incidence represents an area for further study. The additional data we collected on the development of postadmission sepsis, unplanned ICU transfer, 30-day readmission, and mortality can help to inform future research on morbidity and mortality associated with NV-HAP.
Limitations. Although units were selected as matched pairs for their similarity in diagnoses and required care levels, we could not check for cluster control effects on any of the units. The ideal frequency of oral care for NV-HAP prevention is unknown. In this study, accuracy in tracking such frequency was a challenge, as nursing staff didn't always fully document each oral care event and may have recorded self-care that didn't actually occur. Moreover, because oral care was performed by both patients and staff, we could not control for its quality. Hospitals are dynamic institutions and it's not always possible to understand all the potential influencers on interventions and controls. Therefore, it's possible that other hospital initiatives directly or indirectly affected the oral care intervention and NV-HAP outcomes. Lastly, this study was conducted in a single hospital and findings should be interpreted in that context.
Critically ill hospitalized patients are at high risk for NV-HAP and the associated increased morbidity and mortality. NV-HAP prevention is of paramount importance to patient safety. Our study findings suggest that primary source control through improved oral care may be important for NV-HAP prevention. Developing and implementing effective strategies that foster frequent, consistent oral care for all inpatients warrants further study. Moreover, it's not yet known what degree and frequency of oral care is necessary to favorably influence changes in the oral microbiome during acute care hospitalization. Questions regarding ideal oral care frequency, best practices, and consistent implementation would best be addressed through large randomized controlled trials.
1. Magill SS, et al. Changes in prevalence of health care-associated infections in U.S. hospitals. N Engl J Med
2. Cassini A, et al. Burden of six healthcare-associated infections on European population health: estimating incidence-based disability-adjusted life years through a population prevalence-based modelling study. PLoS Med
3. Suetens C, et al. Prevalence of healthcare-associated infections, estimated incidence and composite antimicrobial resistance index in acute care hospitals and long-term care facilities: results from two European point prevalence surveys, 2016 to 2017. Euro Surveill
4. Baker D, Quinn B. Hospital acquired pneumonia
prevention initiative-2: incidence of nonventilator hospital-acquired pneumonia
in the United States. Am J Infect Control
5. Baker D, et al. Sustaining quality improvement: long-term reduction of nonventilator hospital-acquired pneumonia
. J Nurs Care Qual
6. Giuliano KK, et al. The epidemiology of nonventilator hospital-acquired pneumonia
in the United States. Am J Infect Control
7. Micek ST, et al. A case-control study assessing the impact of nonventilated hospital-acquired pneumonia
on patient outcomes. Chest
8. Gomes-Filho IS, et al. Respiratory disease and the role of oral bacteria. J Oral Microbiol
9. Scannapieco FA. The oral microbiome: its role in health and in oral and systemic infections. Clin Microbiol Newsl
10. Scannapieco FA, Shay K. Oral health disparities in older adults: oral bacteria, inflammation, and aspiration pneumonia
. Dent Clin North Am
11. Di Pasquale M, et al. Non-intensive care unit acquired pneumonia
: a new clinical entity. Int J Mol Sci
12. Sopena N, et al. Risk factors for hospital-acquired pneumonia
outside the intensive care unit: a case-control study. Am J Infect Control
13. Heo SM, et al. Genetic relationships between respiratory pathogens isolated from dental plaque and bronchoalveolar lavage fluid from patients in the intensive care unit undergoing mechanical ventilation. Clin Infect Dis
14. Perry SE, et al. The association between oral bacteria, the cough reflex and pneumonia
in patients with acute stroke and suspected dysphagia. J Oral Rehabil
15. Abele-Horn M, et al. Decrease in nosocomial pneumonia
in ventilated patients by selective oropharyngeal decontamination (SOD). Intensive Care Med
16. Kitsios GD, et al. Dysbiosis in the intensive care unit: microbiome science coming to the bedside. J Crit Care
17. Didilescu AC, et al. Respiratory pathogens in dental plaque of hospitalized patients with chronic lung diseases. Clin Oral Investig
18. Gleeson K, et al. Quantitative aspiration during sleep in normal subjects. Chest
19. Huxley EJ, et al. Pharyngeal aspiration in normal adults and patients with depressed consciousness. Am J Med
20. Ewan VC, et al. Dental and microbiological risk factors for hospital-acquired pneumonia
in non-ventilated older patients. PLoS One
21. Quinn B, et al. Basic nursing care to prevent nonventilator hospital-acquired pneumonia
. J Nurs Scholarsh
22. Cassidy MR, et al. The I COUGH multidisciplinary perioperative pulmonary care program: one decade of experience. Jt Comm J Qual Patient Saf
23. McNally E, et al. Oral care clinical trial to reduce non-intensive care unit, hospital-acquired pneumonia
: lessons for future research. J Healthc Qual
24. Munro S, Baker D. Reducing missed oral care opportunities to prevent non-ventilator associated hospital acquired pneumonia
at the Department of Veterans Affairs. Appl Nurs Res
25. Munro S, et al. Implementation and dissemination of a Department of Veterans Affairs oral care initiative to prevent hospital-acquired pneumonia
among nonventilated patients. Nurs Adm Q
26. Pássaro L, et al. Prevention of hospital-acquired pneumonia
in non-ventilated adult patients: a narrative review. Antimicrob Resist Infect Control
27. Lacerna CC, et al. A successful program preventing nonventilator hospital-acquired pneumonia
in a large hospital system. Infect Control Hosp Epidemiol
28. Giuliano KK, Baker D. Sepsis in the context of nonventilator hospital-acquired pneumonia
. Am J Crit Care
29. Sun S, et al. A comprehensive review of effect size reporting and interpreting practices in academic journals in education and psychology. J Educ Psychol
30. Quinn B, Baker DL. Comprehensive oral care helps prevent hospital-acquired nonventilator pneumonia
. Am Nurse Today
31. Tablan OC, et al. Guidelines for preventing health-care–associated pneumonia
, 2003: recommendations of CDC and the Healthcare Infection Control Practices Advisory Committee. MMWR Recomm Rep
32. Boehm LM, et al. Implementation science training and resources for nurses and nurse scientists. J Nurs Scholarsh
33. Powell BJ, et al. Methods to improve the selection and tailoring of implementation strategies. J Behav Health Serv Res
34. Shoemaker SJ. Transforming into a patient-centered medical home: understanding facilitators, barriers and context from a synthesis of implementation studies. Implement Sci
35. Ferguson C, et al. Exploring nursing and allied health perspectives of quality oral care after stroke: a qualitative study. Eur J Cardiovasc Nurs
36. Warren C, et al. A nurse-driven oral care protocol to reduce hospital-acquired pneumonia
. Am J Nurs