Home > RNJ > 2005 > July/August > Evidence-based Clinical Improvement for Mechanically Ventilated Patients

Evidence-based Clinical Improvement for Mechanically Ventilated Patients
Debra C. Hampton, PhD MSN RN • Deborah Griffith, MN RN • Alan Howard, MSN MDiv RN


Bundling or grouping together evidence-based interventions to improve care for the mechanically ventilated patient was piloted by a 10-bed medical-surgical critical care unit of a hospital. The bundled care interventions included: (a) keeping the head of bed elevated at 30 degrees, (b) instituting daily interruption of continuous sedative infusion, (c) assessing readiness to wean using a rapid-wean assessment guide, (d) initiating deep venous thrombosis prophylaxis, and (e) implementing peptic ulcer disease prophylaxis. The interventions were implemented using a plan-do-check-act quality-improvement methodology. Results indicated that the use of bundled interventions for mechanically ventilated patients could decrease average ventilator times and average length of stay with no concomitant increase in reintubations. Average mortality rates and the number of adverse events per 100 patient days also were reduced.

Extensive time and financial resources are spent in completing clinically focused research studies leading to new knowledge that may not be accessible to clinicians who provide patient care. When well-grounded scientific-based evidence is applied in acute care or rehabilitation settings, positive outcomes can occur for patients, clinicians, and even the healthcare institution. Clinical improvements also may result from implementation of evidence obtained from clinical, quality-focused improvements that may not be based on data from randomized clinical trials or other experiment-based evidence (Clemmer, 2003). Regardless of how new evidence is obtained, unless this knowledge is shared with caregivers and applied in the clinical environment, maximum benefit will not be gained. The purpose of this article is to describe how evidence can be combined in a group of clinical interventions directed toward improving the care of mechanically ventilated patients.

Closing the Research/Practice Gap

There are multiple reasons why research-based evidence may take considerable time to be incorporated into practice. As discussed by Berenholtz, Dorman, and Pronovost (2003), clinicians may not be aware of emerging guidelines or research findings, may not agree with their accuracy and utility, or may not have the human or financial resources to implement this information in the practice environment. New information becomes available almost daily, and keeping up-to-date with current knowledge and trends is not easy. In addition, each clinical situation is unique, and applying new techniques or therapies across all situations may not be appropriate.

One solution to the issue of data assimilation and application involves the grouping or bundling of information or measures focused toward a specific disease state or care situation (Berenholtz et al., 2003; Institute for Healthcare Improvement [IHI], 2003; Pronovost, 2002). More significant outcomes may be obtained from implementing multiple changes together, when clinically applicable, than from implementation of a single measure (Berenholtz et al.). Interventions directed toward the care of ventilator patients, whether in an acute care setting or in a rehabilitation setting, are an ideal example of how the care-grouping concept offers clinical benefit.

Bundled Interventions for Mechanically Ventilated Patients

Evidence exists to support the implementation of bundled interventions for mechanically ventilated patients, including elevating the head of the bed to a semirecumbent position, weaning patients from sedation, and follow-up assessment of readiness to wean the patient from the ventilator (Berenholtz et al., 2003; IHI, 2003). The goal is to reduce the number of days patients remain dependent on mechanical ventilation, since mechanical ventilation “is associated with significant morbidity, mortality, and costs of care” (Berenholtz et al., p. 5). Weavind, Shaw, and Feeley (2000) also note that successful ventilator weaning reduces patient morbidity and mortality, along with hospital length of stay (LOS) and associated costs. Pneumonia, which is associated with being on the ventilator, is a common serious complication for critically ill patients that further prolongs hospital LOS and treatment costs (Collard, Saint, & Matthay, 2003). Kollef (1993) also found that patients with ventilator-associated pneumonia (VAP) had a higher mortality rate (37.2%) than those without VAP (8.5%) (p < 0.001).

Recommendations for weaning and discontinuing mechanical ventilation have been formulated (Ely et al., 2001; MacIntyre et al., 2001). One of the most important focus areas in caring for ventilator patients is how patients are positioned. Positioning patients in a semirecumbent position or at a 45-degree angle has been shown to result in decreased ventilator-associated pneumonias, possibly due to a decreased risk for aspiration or gastroesophageal reflux (Collard et al., 2003; Drakulovic et al., 1999; Hixson, Sole, & King, 1998; Kollef, 1993, 1999; Torres et al, 1992; Webster, 1999). Kollef (1993) found that keeping patients supine during the first 24 hours of ventilation resulted in a significant increase in the occurrence of pneumonia; 11.2% of the patients in the semirecumbent position and 34% of the patients in the supine position developed pneumonia (p < 0.001). Kollef also found that the ICU mortality rate for patients in the supine position was significantly higher than for those with the head of bed (HOB) elevated.

Daily interruption of continuous sedation infusions, such as midazolam or propofol, has been shown to be efficacious in the care of mechanically ventilated patients (Ely et al., 2001; Kress, Pohlman, O’Connor, & Hall, 2000). Providing patients with time to wake up and undergo clinical assessment of readiness to wean can promote a more rapid discontinuation of ventilator support (Heffner, 2000; Hong, Mazuski, & Shapiro, 2000; Kress et al., 2000). In a randomized, controlled trial involving 128 patients, Kress et al. found that mechanically ventilated patients with daily interruptions of sedative infusions remained on the ventilator 2.4 days less than patients in the control group (p < 0.001). The median LOS in the ICU in the intervention group was 3.5 days shorter than the LOS in the control group (p = 0.02).

Assessing for readiness to discontinue ventilator support has been demonstrated to result in positive outcomes for the patient and provider. The risk of ventilator-associated pneumonia is decreased with reduced ventilator time (Hixon et al., 1998). Mathews and Mathews (2000) noted, “a patient on mechanical ventilation is 6 to 21 times more likely to develop nosocomial pneumonia than one who is breathing normally. One study suggests that a mechanically ventilated patient’s risk of developing ventilator-associated pneumonia rises 1% per ventilator day” (p. 17). The article further explains that, “between 25% to 50% of all patients on mechanical ventilation acquire one or more infections while on the ventilator. Ventilator-acquired infections are the leading cause of death from nosocomial infections, with reported case fatality rates of 13% to 55%” (p. 17).

Implementing Bundled Ventilator Interventions

Methods for implementing bundled ventilator interventions will vary by setting. The process one hospital used is described here.

A 10-bed medical-surgical critical care unit was selected as the pilot unit. Of the 288 mechanically ventilated patients admitted to the unit, approximately 50% were for respiratory failure, 33% were neurosurgical/neurological, and the remainder were gastrointestinal (GI) surgery, GI medical, and renal. A hospital team implemented a ventilator care bundle in the unit that consisted of five separate interventions: (a) keeping the HOB elevated at 30 degrees, (b) instituting daily interruption of continuous sedative infusion, (c) assessing readiness to wean using a rapid-wean assessment guide, (d) initiating deep venous thrombosis (DVT) prophylaxis, and (e) implementing peptic ulcer disease (PUD) prophylaxis. DVT and PUD prophylaxis will not be discussed here because these interventions do not directly affect ventilator days.

The overall aim of the project, based on recommendations of the Institute for Healthcare Improvement, was to improve outcomes for mechanically ventilated patients as evidenced by (a) decreasing average ventilator time by 50%, (b) decreasing LOS by 20%, and (c) decreasing mortality by 20%, without an increase in the number of reintubations. A compliance goal for ventilator care bundle implementation was set at 90%.

A team was assembled that included a physician champion (pulmonologist/intensivist), the unit’s manager and nurse educator, two clinical nurse specialists (CNSs), a respiratory care specialist, the director of critical care, a clinical pharmacist, and a sponsor from hospital administration. One of the CNSs committed approximately half-time to this project. Two nurse champions from the unit also were engaged to pilot the care bundle. Baseline information was gathered using retrospective analysis of the unit’s previous 12 months of data to establish average ventilator times, average LOS for all patients in the pilot unit, the number of reintubations per month, and the average mortality rate for all patients in the pilot unit.

A plan-do-check-act (PDCA) quality-improvement methodology was chosen for the implementation process to allow for multiple trials using small cycles of change (IHI, n.d.; IHI, 2002; Moen & Provost, 2002). Nursing and respiratory therapy staffs were inserviced on the ventilator care bundle by the unit educator, a CNS, and the respiratory therapist. Physicians were educated about the bundle interventions by the physician champion, the administrative sponsor, and members of the ventilator care bundle team. To help ensure transparent communication with staff and physicians, a bulletin board was placed outside the unit on which was posted information about (a) evidence to support ventilator care interventions, (b) the aim, goals, and desired outcomes of the project, and (c) weekly quality data. Laminated signs also were made that listed the interventions in the ventilator care bundle. A respiratory therapist placed this sign at the HOB for each mechanically ventilated patient receiving the interventions.

The implementation process consisted of applying one intervention from the bundle, piloting the intervention with one nurse champion and one patient, making changes as needed, and repiloting the revised intervention. Once the team and nurse champion felt the intervention was operationalized, it was then released for use by other nurses in the unit. The next intervention was then introduced via a nurse champion using the same small cycle-of-change process. Each intervention was piloted, revised, and introduced to the entire unit over a 3-month period. When the ventilator care bundle was fully implemented within the unit and the team’s focus became one of compliance monitoring, implementation of the bundle into a second critical care unit was initiated using the same small cycle-of-change process.

Baseline data were collected for each unit. Once the ventilator care bundle was completely implemented in a unit, compliance monitoring for that unit commenced, and a weekly graph report card was posted. A monthly report card that included average ventilator time, number of reintubations, average LOS, mortality rate, and monthly ventilator bundle compliance also was posted unit. Compliance monitoring was done 3 times each week, once by the CNS, once by the unit manager, and once by the manager of a different unit. Over a 6-month period, the ventilator care bundle was launched systematically on all units in the critical care division.

The first intervention piloted was keeping the HOB at 30 degrees. Exclusion criteria for HOB intervention included the patient’s refusal, hemodynamic instability, cervical spine instability, and central line placement using a femoral site. To assist in placing the patient at 30 degrees, each bed was clearly marked at 30 degrees using red tape. Eventually, all beds in the critical care division had the 30-degree elevation marked.

The second piloted intervention was daily cessation of continuous sedative infusions. A standing sedation order, based on the collaborative guidelines developed through the Task Force of the Society of Critical Care Medicine and the American Society of Health-System Pharmacists (2002) was used. Patients receiving a continuous sedative infusion for at least 24 hours had the infusion interrupted every day at 0800, and were awakened. If the patient met wake-up criteria, staff could reevaluate the patient’s need for continuous sedative infusion. Based on the nurses’ assessment, in collaboration with the physician, the infusion was either restarted at its existing rate, started at a reduced rate, or discontinued and changed to a PRN sedative dose.

The last intervention introduced was a readiness-to-wean assessment guide. A delicate balance exists between the benefit of aggressively working to discontinue mechanical ventilation and the risk of premature discontinuation (Witta, 1990). Both physiological and psychological criteria for determining patients’ readiness for weaning are important to consider (Blackwood, 2000). A collective task force facilitated by the American College of Chest Physicians, the American Association for Respiratory Care, and the American College of Critical Care Medicine developed a set of evidence-based recommendations for weaning and discontinuing ventilatory support (MacIntyre et al., 2001). The focus of this task force was to find “predictors” of outcomes. Based on the guidelines recommended by this task force, a readiness-to-wean assessment document was developed by the team’s physician champion and the respiratory care specialist. The assessment was printed on a standard physician order/progress note form. Once there was reversal of the underlying cause of respiratory failure, the assessment was implemented every morning by a respiratory therapist.

The various parameters that were measured included: (a) PaO2 (partial pressure of oxygen dissolved in plasma)/FiO2 (fraction of inspired oxygen) or the P/F ratio greater than 150, (b) rapid shallow breathing index (RSBI) less than 100 bpm/l, (c) FiO2 less than .5, (d) positive end-expiratory pressure (PEEP) less than 8 cm H2O, (e) pH greater than 7.33, and (f) mean arterial pressure (MAP) greater than 64 mmHg. A spontaneous breathing trial (SBT) was performed for 30–90 minutes, with a respiratory therapist at the bedside for the first 5 minutes. At the end of the trial, an arterial blood gas (ABG) was obtained. If the pH and PaO2 were acceptable, the physician was contacted for an order to extubate. Additional criteria that were built into the algorithm included inotropes infusing at less than 6 mcg/kg/min, no pressors infusing (e.g., epinephrine, norepinephrine, neosynephrine), and no intra-aortic balloon pump (IABP). Other clinical factors included assessing that the patient was awake and had a normal gag and cough reflex. If the physician gave an order to extubate, an ABG was obtained 30 minutes later.


Implementation of previously discussed bundled care measures for mechanically ventilated patients resulted in positive patient outcomes. The average number of days patients remained on the ventilator dropped from 6.1 to 3.5, which represented a 43% reduction in ventilator days (Figure 1). As would be expected, the patient LOS in the unit also decreased from an average of 4.0 days to an average of 3.4 days (Figure 2). Mortality rate decreased by more than 50%. As ventilator days and mortality decreased, adverse events per 100 patient days were reduced by 73%, from 7 events per 100 patient days to 1.9 events per 100 patient days. At the same time, the number of reintubations remained the same as the rate prior to implementation of the ventilator care bundle.

Implications for Practice

The results of evidence-based, practice-improvement initiatives focusing on the care of mechanically ventilated patients demonstrated positive outcomes from both patient and hospital perspectives. As LOS in the unit went down, the hospital was able to increase admissions to the ICU by 15%.

Patient outcomes improved dramatically. Patients were weaned from the ventilator and extubated much earlier on average than they would have been before implementation of the ventilator care bundle. Unit LOS decreased 15%, with no increase in reintubations or readmissions of patients to the ICU. Mortality and adverse events also dropped significantly, which parallels results obtained by other investigators (Berenholtz et al., 2003).

From the patient’s perspective, the value of reducing ventilator days by almost half is impossible to quantify. Their ability to communicate alone is a priceless benefit, as well as the reduced probability of having an adverse event or developing complications, such as ventilator-associated pneumonia.

Previous studies (Drakulovic et al., 1999; Kollef, 1993; Torres et al., 1992) demonstrating the value of HOB elevation assessed elevation at a semirecumbent position or at a 45-degree angle. The intervention implemented at this institution involved elevating the HOB at 30 degrees. Outcomes from HOB elevation at 30 degrees have not, to date, been widely supported, but the results of the current example demonstrate that positive benefit may result from this elevation height.

The greatest challenge to implementation of the ventilator care bundle was achieving good compliance with HOB elevation. Reasons for noncompliance included: (a) patients slid down in bed, making them uncomfortable and requiring staff to pull them up more frequently, (b) patients did not like having their head elevated because they could not sleep as easily, (c) patients could become more hemodynamically unstable, and (d) patients would experience skin breakdown because they could not lie on their side. Even when reinforced with staff several times weekly, compliance near 100% was not achieved until approximately 1 year after implementation of HOB elevation.

A solution to the issue of how to improve staff compliance with HOB elevation has been offered by Helman et al. (2003). In a study to determine the effect of adding a standard statement to all order sets that reminded staff to elevate the HOB and then following this intervention with formal staff and physician education about the importance of HOB elevation, Helman et al. demonstrated that compliance outcomes improved. Helman et al. found that HOB elevation to a 45-degree angle increased from 3% to 16% 2 months after implementation of the standing order. Further, after an organized staff and physician education program, compliance increased to 29%. Helman et al. also found that after initiating a HOB order and education, the percentage of patients with HOB to 30 degrees or higher increased from 26% at baseline to 83% at 1 month and 72% at 2 months.

We found that ensuring that staff understood the value of HOB elevation, which we demonstrated through sharing evidence from the literature and by communication of positive patient outcomes, removed the greatest roadblock to success. Added to this, consistent monitoring and the reinforcement of clear expectations, along with sensitivity in working with staff made the difference in compliance success. Education of ancillary support staff and family also was critical, so that they would not lower the patient’s head.


These findings support the value of implementing evidence-based improvements in the care of mechanically ventilated patients, as addressed by Berenholtz et al. (2003) and others (IHI, 2003). Spreading the implementation of bundled ventilator care measures beyond the ICU into rehabilitation units may result in reduction in ventilator days and LOS, as demonstrated by positive outcomes achieved to date in a pulmonary transitional care unit that admits patients with complex respiratory diagnoses. As noted by the IHI, “there is no dazzling new research or technical wizardry behind the ventilator bundle steps” (2003, p. 3). But the outcomes from the implementation of combined evidence-based practices are making a positive difference for many ventilator patients.

About the Authors

Debra Hampton, PhD MSN RN, is director of nursing, Critical Care/Telemetry/Transitional & Interventional Care at Saint Joseph Healthcare in Lexington, KY, and a part-time faculty member of the University of Kentucky College of Nursing in Lexington.

Deborah Griffith, MN RN, is a critical care clinical nurse specialist at Saint Joseph Hospital in Lexington, KY.

Alan Howard, MSN RN M. Div., is assistant director, Emergency Department, The Reading Hospital and Medical Center, Reading, PA.

Address correspondence to Debra Hampton, PhD MSN RN, 2244 Savannah Lane, Lexington, KY 40513, or via e-mail to DHampton@sjhlex.org.


Berenholtz, S. M., Dorman, T., & Pronovost, P. (2003). Improving quality and safety in the ICU. Contemporary Critical Care, 1(1), 1–10.

Blackwood, B. (2000). The art and science of predicting patient readiness for weaning from mechanical ventilation. International Journal of Nursing Studies, 37, 145–151.

Clemmer, T. P. (2003, June). A multi-pronged approach to reducing ICU infections. Paper presented at the Institute of Health Care Improvement Spring Learning Session, Boston, MA.

Collard, H. R., Saint, S., & Matthay, M. A. (2003). Prevention of ventilator-associated pneumonia: An evidence-based systematic review. Annals of Internal Medicine, 138, 494–501.

Drakulovic, M. B., Torres, A., Bauer, T. T., Nicolas, J. M., Nogue, S., & Ferrer, M. (1999). Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: A randomized trial. The Lancet, 354, 1851–1858.

Ely, E. W., Meade, M. O., Haponik, E. F., Kollef, M. H., Cook, D. J., Guyatt, G. H. et al. (2001). Mechanical ventilator weaning protocols driven by non-physician health care professionals: Evidence-based clinical practice guidelines. Chest, 120, 454S–463S.

Heffner, J. E. (2000). A wake-up call in the intensive care unit. [Electronic version]. The New England Journal of Medicine, 33, 1520–1522.

Helman, D. L., Sherner, J. H., Fitzpatrick, T. M., Callender, M. E., & Shorr, A. F. (2003). Effect of standardized orders and provider education on head-of-bed positioning in mechanically ventilated patients. Critical Care Medicine, 31, 2285–2290.

Hixon, S., Sole, M. L., & King, T. (1998). Nursing strategies to prevent ventilator-associated pneumonia. [Electronic version]. AACN Clinical Issues, 9(1), 76–90.

Hong, J. J., Mazuski, J. E., & Shapiro, M. C. (2000). Daily interruption of sedative infusions in critically ill patients. [Electroinc version]. The New England Journal of Medicine, 343, 814–815.

Institute for Healthcare Improvement. (n.d.). Quality improvement resources: A model for accelerating improvement. Retrieved November 3, 2003, from http://www.ihi.org/resources/qi/index.asp

Institute for Healthcare Improvement. (2003, October). Improvement tip: Bundle up. IHI Continuous Improvement Newsletter. Retrieved October 27, 2003, from http://www.ihi.org/resources/qi/qitips/ci1003tip.asp.

Institute for Healthcare Improvement. (2002, June). Improvement tip: Test changes on a small scale. IHI Continuous Improvement Newsletter. Retrieved November 3, 2003, from http://www.ihi.org/resources/qi/qitips/ci0602tip.asp.

Kollef, M. H. (1999). The prevention of ventilator-associated pneumonia. The New England Journal of Medicine, 340, 627–634.

Kollef, M. H. (1993). Ventilator-associated pneumonia. JAMA: The Journal of the American Medical Association, 270, 1965–1970.

Kress, J. P., Pohlman, A. S., O’Connor, M. F., & Hall, J. B. (2000). Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. The New England Journal of Medicine, 342, 1471–1477.

MacIntyre, N. R., Cook, D. J., Ely. E. W., Jr., Epstein, S. K., Fink, J. B., Heffner, J. E. et al. (2001). Evidence-based guidelines for weaning and discontinuing ventilatory support. Chest, 120, 375S–395S.

Mathews, P. J., & Mathews, L. M. (2000). Reducing the risks of ventilator-associated infections. Dimensions of Critical Care Nursing, 19(1), 17–21.

Moen, R., & Provost, L. (2002, October). Model for improvement: Aims and measures. Paper presented at the Institute for Health Care Improvement Fall Learning Session, Boston, MA.

Pronovost, P. (2002). Care Bundles. Institute for Health Care Improvement Fall Learning Session. October 10-11, 2002. Boston, MA.

Society of Critical Care Medicine & American Society of Health-System Pharmacists. (2002). Clinical practice guidelines for the sustained use of sedatives and analgesics in the critically ill adult. American Journal of Health-System Pharmacists, 59, 150–178.

Torres, A., Serra-Batlles, J., Ros, E., Piera, C., Bellacasa, J., Cobos, A. et al. (1992). Pulmonary aspiration of gastric contents in patients receiving mechanical ventilation: The effect of body position. Annals of Internal Medicine, 116, 540–543.

Weavind, L., Shaw, A. D., & Freeley, T. W. (2000). Monitoring ventilator weaning: Predictors of success. Journal of Clinical Monitoring and Computing, 16, 409–416.

Webster, N. R. (1999). Importance of position in which patients are nursed in intensive care units. The Lancet, 354, 1835–1836.

Witta, K. (1990). New techniques for weaning difficult patients from mechanical ventilation. AACN Clinical Issues, 1, 260–266.