It's been well understood for the past two decades that medical errors and hospital-acquired infections (HAIs) have a significant negative impact on the health of Americans and are highly costly to treat. Hospital-acquired pneumonia (HAP) accounts for a large part of the health and cost burden. A systematic review of HAIs that examined the literature from 1986 through 2013 found that ventilator-associated pneumonia (VAP), a subset of HAP and the leading cause of death in critically ill patients, was the second most costly of the top five HAIs.1 The researchers estimated that the average case of VAP costs more than $40,000. They estimated the total annual cost of the top five HAIs (central line–associated bloodstream infection [CLABSI], VAP, surgical site infection, Clostridium difficile infection, and catheter-associated urinary tract infection [CAUTI]) to be $9.8 billion, with VAP costing nearly a third (31.6%) of the annual total, or approximately $3.1 billion.
Both nonventilator hospital-acquired pneumonia (NV-HAP) and VAP remain prevalent and problematic.2 In 2012 the Centers for Disease Control and Prevention (CDC) reported an overall critical care mean ventilator use of 0.24 to 0.45 per 100 patient days, with VAP occurring on critical care units at a mean rate of 0.2 to 4.4 per 1,000 ventilator days.3 In 2013, the CDC updated the definition of reportable VAP as “ventilator-associated events” (VAEs), which include possible or probable VAP, ventilator-associated conditions, and infection-related ventilator-associated complications.4
While many factors likely contribute to the development of HAP and VAEs, evidence suggests two primary connections between oral care and pneumonia; that is, the presence of bacteria and the pathogenesis of periodontitis. First, many bacterial species can be found in the oral cavity on the mucosa and in the dental plaque, the biofilm found on the surface of the tooth where these bacteria colonize.5 Potentially, the mouth harbors more than 600 varieties of microorganisms, which colonize on the tooth surface and in the oral mucosa.6 Streptococcus pneumoniae, Haemophilus influenzae, Staphylococcus aureus, and Enterobacter species are a few examples of potentially harmful bacteria that may colonize in the oral cavity.7 Gram-negative bacteria triple on the tooth surface in a matter of three to six hours.8
The second connection between oral care and pneumonia is related to the pathogenesis of periodontitis, in which an inflammatory process results in the release of enzymes and cytokines and facilitates the adherence of harmful bacteria to teeth and mucosa, compounding the problem.9 Coupled with the lack of oral care, patients’ microaspiration of colonized bacteria into the upper airway, where bacteria can migrate into the lungs, may lead to development of pneumonia.10
DESCRIPTION OF THE PROBLEM
After attending an internal seminar led by a surgeon who presented opportunities for improvement in postoperative pulmonary outcomes, two bedside surgical unit nurses (two of us, MKM and BW) applied for an evidence-based practice (EBP) fellowship to examine potential nursing interventions (such as incentive spirometry and oral care) to improve these outcomes. Through this fellowship, the nurses collaborated with a critical care clinical nurse specialist (CNS) (one of us, CW) and a nurse researcher (DS). A preliminary survey of evidence, observed inconsistencies in delivery of oral care, and staff perceptions of patient dissatisfaction with oral care products led to a focused examination of evidence related to oral care and pulmonary outcomes in order to develop a feasible evidence-based protocol for bedside nurses.
The hospital's previous oral care products were lacking; for example, they didn't contain any agents such as baking soda, didn't include suction toothbrushes for at-risk patients, and weren't bundled for ease of use. The first phase of this project resulted in the pilot implementation by a quality improvement (QI) team of a standardized oral care protocol including standardized products on a single general surgical unit. After successful implementation of the pilot, we refined the protocol and expanded the size of the team, the scope of the outcome measurement (to include HAP), and the practice change and product implementation. The hospital-wide implementation and evaluation of the expanded nurse-driven oral care protocol and use of the new oral care kits are the focus of this report.
LITERATURE REVIEW AND SYNTHESIS OF THE EVIDENCE
First, we conducted a review of the literature to answer a clinical question drafted in PICO format—a format that specifies the population of interest, intervention of interest, comparison intervention, and outcome measure. The question was: For adult inpatients in acute care (P), would the implementation of a nurse-driven oral care protocol (I) compared with no protocol (C) reduce the incidence of HAP (O)? A search of the PubMed, CINAHL, and Cochrane databases yielded 25 studies. Eighteen studies were excluded because they did not test oral care interventions or were not systematic reviews of intervention studies. We evaluated and synthesized the seven remaining studies, ranging from meta-analyses to retrospective cohort studies, to inform a standardized oral care program and nursing practice change. The studies were synthesized and critiqued based on Melnyk and Fineout-Overholt's hierarchy of evidence.11 The studies included two meta-analyses (level 1), three controlled trials (level 3) or cohort studies (level 4), and two quasi-experimental comparative studies (level 3).
In 2003, the CDC issued guidelines recommending the development and implementation of a comprehensive oral care protocol for in-hospital patients. Subsequent evidence suggested that standardized oral care is, in fact, effective in improving pulmonary outcomes. In a 2015 meta-analysis by Kaneoka and colleagues, for example, oral care was associated with a reduced risk of developing pneumonia and a decreased mortality risk from pneumonia in hospitalized patients (though the authors noted that the effect “should be interpreted with caution due to risk of bias in the included trials”).12 Other studies of standardized or intensified oral care regimens have shown reductions in acquired pneumonia rates in acute care2, 10, 13, 14 and long-term care settings.15 Notably, Quinn and colleagues reported a 37% decrease in NV-HAP and cost avoidance of an estimated $1.6 million over 12 months on implementation of a standardized oral care program.2
We derived specific components of a comprehensive oral care protocol from several studies; however, our review of the literature shows there is no single high-quality standard of care. For example, one study that examined a range of brushing times and their effect on plaque removal recommended that oral health care professionals should encourage patients to brush for 120 seconds two to three times daily; most estimates of actual brushing times range from approximately 30 to 60 seconds.16 We adopted the 120-second brushing time for our protocol, every four hours (six times per day) for patients receiving mechanical ventilation and every six hours (four times per day) for nonventilated patients, as suggested by Quinn and colleagues.2 High-quality oral care also includes the use of baking soda toothpaste, antiseptic mouth rinse, and mouth moisturizer.2, 12 Evidence also suggests a high-quality oral care protocol should include educational information for patients2 and early dysphagia screening.14 We also drew on the oral care protocol developed by Weitzel and colleagues to prevent nosocomial pneumonia.10
We synthesized the studies and concluded that an evidence-based oral care program would need to take a comprehensive approach, including improved, risk-stratified oral care products bundled for ease of storage, access, and use and a protocol for improved oral care delivery (provided four to six times daily, depending on risk level), routine assessment, and patient education. Staff education and program evaluation would also be an important part of the implementation and evaluation of the QI project. The project team, including the two bedside nurses and critical care CNS, worked together to identify products and develop the oral care protocol.
The purpose of this QI project was to evaluate the hospital-wide implementation of a nurse-driven oral care protocol, including improved oral care product kits and a collaboratively developed educational information sheet for patients and family caregivers.
Setting. This project was conducted in all adult in-patient care areas at a level 1 trauma hospital in mid-Michigan and included all adult inpatient nurses and patient care technicians (PCTs). Our nonprofit hospital has more than 650 licensed beds and is a Magnet-certified center for nursing excellence.
Description of the oral care practice change. The evidence-based oral care practice change included a standardized, evidence-based oral care protocol (see Figure 1) that provided guidance in determining the level of oral care required for a patient in each of three categories: ventilated, at risk, and short-term care. Patients not at risk for aspiration received a standard manufacturer-packaged short-term oral care kit to be used four times daily that included an ergonomically appropriate toothbrush with higher-quality bristles, an alcohol-free antiseptic mouth rinse, a baking soda toothpaste, mouth moisturizer, and oral care swabs with baking soda. Patients who weren't ventilated but were determined to be at risk for aspiration received the manufacturer-packaged at-risk oral care kit, including the following components: a suction toothbrush, an alcohol-free antiseptic mouth rinse, and mouth moisturizer. Oral care for at-risk patients was also to be delivered four times daily. The ventilator oral care kit was already in use in the facility and included a suction toothbrush, chlorhexidine mouth rinse to be used once every 12 hours, antiseptic mouth rinse, and mouth moisturizer. The oral care protocol specifies that the ventilator oral care kit is to be used six times daily.
QI implementation procedures. The critical care CNS provided clinical nurse education on the new QI protocols in inpatient RN education sessions 12 times during fall 2015. Content included the pathogenesis of periodontitis, including microaspirations of colonized oral bacteria, a brief review of the literature on oral hygiene and bacterial colonization, and a review of the newly drafted protocol and expectations regarding the type and frequency of assessments nurses were to perform. A total of 1,131 clinical nurses completed the educational component (approximately 87% participation). The critical care CNS provided similar role-specific education to physical therapists, occupational therapists, and speech and language therapists on two separate occasions in fall 2015; she also provided training (both didactic and hands on) to PCTs in multiple concurrent sessions. Finally, an educator representing the oral care product manufacturer provided additional in-service programs on the new oral care products to day- and night-shift caregivers at the unit level during the two-week project rollout.
The QI team developed a patient education information sheet in collaboration with the corporate funder, Delta Dental Foundation; the hospital's patient education committee approved it, and the team incorporated it in the new oral care protocol and care-product kits (see Figure 2). In practice, the bedside nurses reviewed the information sheets with patients to reinforce the importance of consistent oral care. Key points in the patient education sheet included the importance of reducing colonized bacteria through oral hygiene interventions four to six times daily in order to prevent HAP.
Finally, QI team members worked with the hospital's information technology (IT) department to modify the hospital's electronic health record (EHR) to support the use and documentation of the oral care protocol and products. These modifications provided decision support and a visual display of the step-by-step procedures in the EHR to cue practice and allow more detailed documentation. These documentation changes, in combination with product use and hospital length of stay and discharge reports, allowed us to monitor nurses’ and PCTs’ adherence to the protocol.
Evaluation procedures. This QI project used a quasi-experimental pretest–posttest evaluation plan. Several team members measured both process and outcome indicators to evaluate the impact of the practice change. The primary process indicator was nurses’ adherence to the practice change. The primary outcome indicator was the incidence of NV-HAP and VAP. Patient chart review procedures were consistent with evidence-based QI measures, as no patient-specific data were collected. The hospital's institutional review board granted a nonresearch determination for the implementation and evaluation of the nurse-driven oral care protocol.
Nurse and PCT adherence. Several members of the QI team measured staff members’ adherence to the protocol by examining EHR documentation of oral care interventions and product use reports and compared them with estimations of oral care interventions based on the number of patients on a given unit and their lengths of stay. The IT department generated the product use reports monthly. The QI team calculated the expected product use per unit by multiplying the number of patient days by the frequency of protocol-specified oral care (four times per day or six times per day, for example) on each unit. The actual use was determined by measuring unit supply orders and documented oral care interventions. The overall adherence (per unit and hospital-wide) was calculated as a percentage of actual use compared with expected use (that is, actual use divided by expected use multiplied by 100).
NV-HAP and VAP outcomes. Key HAP patient outcome indicators monitored in this project included NV-HAP and VAP incidence. A report was generated in the EHR using a list of identified pneumonia-specific International Classification of Diseases (ICD) 9 and ICD 10 codes (see Table 1) for patients with pneumonia diagnoses not present on admission between two seven-month periods: November 2014 to May 2015 and November 2015 to May 2016, which were the baseline and intervention periods of the QI project, respectively. The charts were then reviewed to determine whether patients met the inclusion criteria for NV-HAP according to the CDC's PNU1 criteria algorithm.17 Key elements of the CDC criteria included a chest X-ray with correlating clinical signs and symptoms—specifically, a positive chest X-ray plus fever or elevated white blood cell count or change in mental status, in addition to two signs and symptoms (such as onset of or worsening sputum, onset of or worsening cough, dyspnea, tachypnea, rales or bronchial breath sounds, worsening gas exchange, and increased oxygen requirements) that appear following the second day after admission in a seven-day period.
Data analysis procedures. Chart audit data were transferred into an Excel spreadsheet for analysis. Data analyses were conducted using SPSS Version 24 (IBM Corp.). Summary statistics were calculated in order to describe key variables, and χ2 analyses were used to compare selected indicators before and after implementation. Statistical significance was set at P < 0.05.
Nurse and PCT adherence to protocol. Staff adherence to protocol varied across nursing units (36% to 100% per month) with an overall adherence of 76%. No obvious trends emerged when comparing adherence by unit type (medical versus surgical, for example) or unit acuity (critical care versus acute care).
HAP incidence. NV-HAP. The ICD 9 and ICD 10 pneumonia report yielded 202 patient charts in the baseline group and 215 patient charts in the intervention group for review. Using the CDC's PNU1 criteria algorithm, 52 cases of NV-HAP were identified in the baseline period, compared with 26 cases in the intervention period, a statistically significant difference (χ2 = 12.8, df = 1, P < 0.001). There were 2.84 infections per 1,000 discharges in the baseline group compared with 1.41 infections per 1,000 discharges in the intervention group. There were also 20 deaths in the baseline group compared with four in the intervention group (χ2 = 4.33, df = 1, P = 0.037). (See Table 2.)
VAP incidence. Among patients receiving mechanical ventilation, 56 cases of VAEs with 12 cases of VAP were identified in the baseline group compared with 49 VAEs with three cases of VAP in the intervention group. Infection rates were also calculated, yielding 12.53 VAEs and 2.87 cases of VAP per 1,000 ventilator days in the baseline group and 14.29 VAEs and 1.26 cases of VAP per 1,000 ventilator days in the intervention group.
Hospital-wide implementation of a multicomponent oral care bundle, including a decision-making algorithm, tailored oral care product kits, and embedded patient education, resulted in a significant improvement in pneumonia outcomes. The results of this QI project included a 50% reduction in the incidence of NV-HAP despite an increase in patient days and discharges. In addition, an estimated 16 deaths and $1.04 million in costs were avoided through prevention of NV-HAP over a seven-month evaluation period. These findings are consistent with those of other practice change initiatives, such as that of Quinn and colleagues, who reported a 37% reduction in NV-HAP with similar cost savings over a similar time period.2
While the overall adherence to protocol by nursing staff was respectable (76%), staff adherence did vary across units, suggesting the need for ongoing, targeted feedback and additional staff education. Notably, since the QI project, several nursing units have chosen oral care as their quarterly peer review project to further self-monitor this important patient intervention protocol. Assessing adherence was challenging and involved triangulating several sources of data, including monthly product use by unit (unit supply reports), patient days per unit, and nursing documentation of oral care. In reviewing the literature, we were not able to identify other projects that reported adherence to protocol in this manner.
Limitations. Several limitations and lessons learned were noted as a result of the hospital-wide implementation of this practice change. To facilitate major product change in an organization with 650-plus beds requires complex project management. More detailed gap analysis would have been helpful, as the availability of equipment, such as bedside suction devices, and other resources varies across units and buildings. Prior to this practice change, we had also determined that our PCTs had little to no education on the use of oral suction. Although nurses and PCTs understand the value of comprehensive oral care, a gap still exists between knowledge and practice. Ongoing education and auditing of oral care interventions are necessary to help minimize missed opportunities to provide nursing care. As demonstrated in this project, staff and patient education on the importance of oral care interventions does have a positive effect on practice. There were also difficulties in obtaining key patient outcome data for NV-HAP, because it is not a reportable condition and, therefore, not regularly tracked by the hospital's infection prevention department. The QI team worked with the IT department to develop the reports and then evaluated each patient's chart to determine whether a given patient met the CDC's PNU1 criteria. Determination of readmission rates for the groups identified was considered outside the scope of this project and, therefore, the rates were not obtained.
Our findings highlight the importance of tracking these outcomes through an automated or regularly managed method rather than by reviewing charts individually, in addition to clinical work. Health care organizations would do well to track and report the incidence of NV-HAP, as we currently track and report other nursing quality indicators, such as falls, CLABSIs, and CAUTIs, to improve patient safety and outcomes.
Key strengths. Despite these challenges and lessons learned, this project has several key strengths. For example, use of the CDC PNU1 criteria to validate the inclusion of each pneumonia case as an occurrence of NV-HAP or VAP increases the validity of our results. An additional strength of this QI project was its origin as a nurse-developed and nurse-driven protocol and intervention. This project grew out of an EBP fellowship initiative on a single medical–surgical nursing unit led by two bedside nurses. The successful pilot test expanded under the direction of the critical care CNS and included an expanded interdisciplinary team of bedside nurses, a nurse research consultant, physicians, an infection preventionist, a speech/language therapist, an IT analyst, a nurse educator, the philanthropy director of the health system's foundation, and support from the Delta Dental Foundation, an affiliate of Delta Dental of Michigan, Ohio, and Indiana.
Implications. The implementation of a new, standardized, evidence-based oral care protocol was a nurse-initiated, nurse-driven effort to transform a relatively simple, overlooked nursing intervention into a high-priority, high-impact daily intervention tailored to the needs of hospitalized patients. Nurses improved pneumonia outcomes by providing oral health interventions and targeted patient education for all adult patients admitted to the hospital, which reduced overall incidence of HAP, hospital costs, and patient mortality. These successful results occurred because nurses were empowered to lead the way in improving the monitoring and implementation of effective oral care interventions for the prevention of NV-HAP.