Kazaure HS, Martin M, Yoon JK, Wren SM. Long-term Results of a Postoperative Pneumonia Prevention Program for the Inpatient Surgical Ward. JAMA Surg. 2014;149(9):914-918. doi:10.1001/jamasurg.2014.1216
Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Pneumonia is the third most common complication in postoperative patients and is associated with significant morbidity and high cost of care. Prevention has focused primarily on mechanically ventilated patients. This study outlines the results of the longest-running postoperative pneumonia prevention program for nonmechanically ventilated patients, to our knowledge.
To present long-term results (2008-2012) of a standardized postoperative ward-acquired pneumonia prevention program introduced in 2007 on the surgical ward of our hospital and compare our postintervention pneumonia rates with those captured in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP). We also estimate the cost savings attributable to the pneumonia prevention program.
Design, Setting, and Participants
Retrospective cohort study at a university-affiliated Veterans Affairs hospital of all noncardiac surgical patients with ward-acquired postoperative pneumonia.
A previously described standardized postoperative pneumonia prevention program for patients on the surgical ward.
Main Outcome and Measure
Ward-acquired postoperative pneumonia.
Between 2008 and 2012, there were 18 cases of postoperative pneumonia among 4099 at-risk patients hospitalized on the surgical ward, yielding a case rate of 0.44%. This represents a 43.6% decrease from our preintervention rate (0.78%) (P = .01). The pneumonia rates in all years were lower than the preintervention rate (0.25%, 0.50%, 0.58%, 0.68%, and 0.13% in 2008-2012, respectively). The overall pneumonia rate in ACS-NSQIP was 2.56% (14 033 cases of pneumonia among 547 571 at-risk patients), which is 582% higher than the postintervention rate at our ward. Using a national average of $46 400 in attributable health care cost of postoperative pneumonia and a benchmark of a 43.6% decrease in pneumonia rate achieved at our facility over the 5-year study period, a similar percentage of decrease in pneumonia occurrence at ACS-NSQIP hospitals would represent approximately 6118 prevented pneumonia cases and a cost savings of more than $280 million.
Conclusions and Relevance
The standardized pneumonia prevention program achieved substantial and sustained reduction in postoperative pneumonia incidence on our surgical ward; its wider adoption could improve postoperative outcomes and reduce overall health care costs.
Quiz Ref IDPneumonia is a common nosocomial infection that accounts for approximately 15% of all hospital-acquired infections1 and 2.7% to 3.4% of complications among surgical patients.2 Nosocomial pneumonia is associated with high morbidity, mortality, and cost of care.2,3 It is among index complications used to assess readmission rates for hospital profiling and reimbursement as endorsed in recent health care reform legislation.4 As expected, increased attention to the improvement of quality of patient care with emphasis on infection control has spurred the development of several guidelines for nosocomial pneumonia prevention.
Guidelines and interventions aimed at reducing the postoperative patient's risk for pneumonia continue to focus on the critical care population, specifically those on a mechanical ventilator, ie, ventilator-associated pneumonia. Prior to implementation of our pneumonia prevention program for patients on the surgical ward in 2007,5 our hospital had a higher rate of postoperative pneumonia when compared with other Veteran’s Affairs (VA) hospitals; specifically, the surgical ward averaged 8.7 cases of pneumonia per year. After successful implementation of the standardized, bundled pneumonia prevention program, we reported an 81% decrease in ward-acquired pneumonia occurrence in the 18 months following the introduction of the prevention program.
Since our publication5 in 2010, only a single study6 has reported results of a prevention program for postoperative pneumonia in non–intensive care unit settings. In their description of the I COUGH program for preventing pulmonary complications among surgical patients at their institution, Cassidy and colleagues6 did audits on interventions such as head-of-bed elevation, out-of-bed status, and incentive spirometer use. Echoing our first experience after implementation of the program, Cassidy and colleagues showed a 38.5% decrease in the postoperative pneumonia rate after 1 year of introducing the multidisciplinary program.
Considering the scarcity of a standardized approach to pneumonia prevention in non–critical care settings, it is not unexpected that there are no studies, to our knowledge, describing the long-term impact of a standardized postoperative pneumonia prevention program among surgical patients receiving care in non–critical care settings. To evaluate the long-term effectiveness of our program, focused entirely on patients hospitalized on the surgical ward, we present the results of this program for a period of 5 years since implementation. In recognition of the importance of demonstrating sustained reduction of postoperative pneumonia incidence in our surgical ward to the potential widespread adoption of our pneumonia prevention program, we compare our postintervention rates with pneumonia rates captured in a multi-institutional database.
The pneumonia prevention quality improvement task force at VA Palo Alto Health Care System was formed in December 2006. After education and orientation of relevant hospital personnel, evidence-based ward-acquired pneumonia prevention strategies were implemented in April 2007. The full details of this intervention program have been described in a prior publication.5 Briefly, the program consisted of the following 8 steps:
Initial and ongoing education of all surgical ward nursing staff about their role in pneumonia prevention.
Coughing and deep-breathing exercises with incentive spirometer.
Twice-daily oral hygiene with chlorhexidine.
Ambulation with good pain control.
Head-of-bed elevation to at least 30° and sitting up for all meals (“up to eat”).
Quarterly discussion of the progress of the program and results for nursing staff.
Pneumonia bundle documentation in the nursing documentation.
Automated computerized physician pneumonia prevention order set in physician order entry system.
No new preventive measures were introduced during the 5-year study period. The task force during this period consisted of a physician (assistant chief of the Surgical Service), VA Surgical Quality Improvement Program (VA-SQIP) nurse, and the surgical ward nurse manager who reviewed pneumonia statistics on an ongoing monthly basis.
Data used in this study are deidentified and part of a quality improvement study that was deemed exempt by Stanford University Institutional Review Board as not meeting the federal definition of research or clinical investigation. Prospective data collection followed by retrospective review were performed of all inpatient pneumonia cases that occurred among patients hospitalized on the surgical ward documented in the VA-SQIP database at VA Palo Alto over a 5-year period (2008-2012). Data for 2007 were excluded because this represents a transition period. As in the preintervention period,5 inclusion criteria for the study cohort were all noncardiac surgical patients with an inpatient admission to the surgical ward included in the VA-SQIP program; each patient was scored for presence or absence of ward-acquired pneumonia. Exclusion criteria were all outpatient operations, admissions to the observation 23-hour stay unit, admissions to a nonsurgical ward, and patients who were diagnosed with pneumonia in an intensive care unit. The definition of pneumonia, as defined by VA-SQIP, was unchanged. Patients must meet 1 of 2 criteria: (1) new onset of purulent sputum or change in character of sputum; isolation of an organism from blood culture; or isolation of pathogen from specimen obtained by transtracheal aspirate, bronchial brushing, or biopsy or (2) histopathologic evidence of pneumonia. In addition, they must meet 1 of the following 2 criteria: (1) rales or dullness to percussion on physical examination of the chest or (2) chest radiograph that demonstrates a new or progressive infiltrate, consolidation, cavitation, or pleural effusion.
We also used data collected from up to 350 US hospitals participating in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) that were included in the Participant Use Data Files (2008-2012). The ACS-NSQIP and its method of data collection are well described.7 The ACS-NSQIP excludes all data from VA hospitals and uses a data collection system identical to VA-SQIP. The definition of postoperative pneumonia in the ACS-NSQIP is the same as that of the VA-SQIP. To allow adequate comparison with our data, the ACS-NSQIP cohort included in this study comprised all male patients who underwent inpatient, noncardiac surgery captured in the database; those with preoperative pneumonia and those cases that may represent ventilator-associated pneumonia (patients with a diagnosis of ventilator dependence or ventilator use, postoperative reintubation, failure to wean, or preoperative or postoperative coma) were excluded from the cohort.
Documented cases of pneumonia in each year were divided by the total number of patients at risk in that period and expressed as an annual case rate. The χ2 test of goodness of fit was used to determine whether the decrease in pneumonia rate in the postintervention period was significant compared with the preintervention rate. The overall pneumonia rate over the 5-year period (2008-2012) at VA Palo Alto was computed and compared with the overall rate in ACS-NSQIP using the t test. A P value <.05 was considered significant.
Data regarding cost of care are not reported in either VA-SQIP or ACS-NSQIP. Therefore, to glean insight into the financial implications of the pneumonia prevention program, we estimated the attributable cost saved per year following introduction of the pneumonia prevention program using a national average of $46 4008 in health care cost attributable to 1 case of postoperative pneumonia.
Data analyses and management were performed using SPSS statistical software for Windows, version 19.0 (IBM).
Between 2008 and 2012, within the study cohort there were 18 cases of ward-acquired pneumonia among 4099 at-risk patients hospitalized on the VA Palo Alto surgical ward, yielding an average case rate of 0.44%. When compared with a preintervention rate of 0.78%, this represents a 43.6% overall decrease in pneumonia diagnosis among surgical patients hospitalized in the ward (P = .01). The pneumonia case rates in all postimplementation years were lower than the preintervention rate (Table).
The overall pneumonia case rate in ACS-NSQIP was 2.56% (14 033 cases of pneumonia among 547 571 at-risk patients) over the 5-year study period. This was 582% higher than the average postintervention rate at our ward; the difference is statistically significant (P < .001). The annual postoperative pneumonia rates in ACS-NSQIP were also consistently higher for each year compared with postintervention rates at our surgical ward (Table).
Quiz Ref IDThe average annual pneumonia rate in the surgical ward at VA Palo Alto decreased from 8.7 cases per year in the preintervention period to an average of 3.6 cases per year during the 5-year postintervention period; in other words, approximately 5 cases of pneumonia were prevented per year.Quiz Ref IDUsing a national average of $46 400 in health care cost attributable to 1 case of postoperative pneumonia, this represents approximately $232 000 in attributable cost of pneumonia saved per year since the introduction of the pneumonia prevention program at our surgical ward. Using the same national average of $46 400 in attributable health care cost of pneumonia and a benchmark of a 43.6% decrease in pneumonia rate achieved at our facility over the 5-year study period, a similar percentage of decrease in pneumonia occurrence at ACS-NSQIP hospitals would represent 6118 prevented pneumonia cases and a cost savings of $283 875 200.
To our knowledge, this is the first presentation of long-term results of a standardized postoperative pneumonia prevention program for surgical inpatients who are not on a mechanical ventilator. Since 2007, we prioritized the eradication of pneumonia among surgical patients hospitalized on the surgical ward at VA Palo Alto. The introduction of the program and concerted effort to maintain compliance with the program since its inception 5 years ago have led to a substantial and sustained decrease in the pneumonia rate. Indeed, there was only 1 case of pneumonia in 2012 among more than 750 postoperative patients hospitalized in the VA Palo Alto surgical ward; this indicates that the goal of zero cases of pneumonia is achievable.
The postoperative pneumonia rate in ACS-NSQIP is, on average, similar to that of the VA Palo Alto facility prior to inception of the pneumonia prevention program.5 In all postintervention years studied, the pneumonia rate in ACS-NSQIP hospitals was at least 3 times the rate documented in the VA Palo Alto surgical ward, highlighting an avenue for targeted improvement. We estimated that more than 6100 cases of pneumonia could be prevented with a 43.6% reduction in pneumonia rate among ACS-NSQIP hospitals. We showed that the economic implication of decreasing postoperative pneumonia occurrence is tremendous. Our analysis demonstrated that an average 43.6% reduction in postoperative pneumonia incidence among surgical patients in ACS-NSQIP hospitals would correlate with health care cost savings of more than $280 million.
In part because of renewed focus on patient safety as well as recent health care legislation and Medicare reimbursement policies aimed at decreasing morbidity and expensive health care expenditures, considerable effort is being expended across institutions in the country to reduce the occurrence and hospital readmission secondary to a diagnosis of pneumonia. The results of this study demonstrate that the occurrence of non–ventilator-associated pneumonia among surgical patients can be prevented by adhering to bundles of infection prevention measures. Although there are no established guidelines for pneumonia prevention for nonmechanically ventilated patients, we believe that adoption of ours or similar pneumonia prevention programs at other institutions would lead to considerable reduction in postoperative pneumonia rates. Quiz Ref IDThe use of an electronic medical record system now ubiquitous at hospitals would facilitate the standardized application of elements of our pneumonia prevention program at other hospitals via the inclusion of order sets for the pneumonia prevention bundle in a patient’s postoperative care orders that are placed electronically. We emphasize, however, that from our experience, the commitment of all members of the multidisciplinary pneumonia prevention task force is key to the success of the program.
The limitations of this study include the use of database abstraction methods to select patients with pneumonia. Errors are possible; however, both the VA-SQIP and ACS-NSQIP data abstraction methods are well validated.7,9 Moreover, the definition of pneumonia in the 2 databases is identical. Trend analysis of pneumonia occurrence over time at our surgical ward was precluded by the overall small number of cases of pneumonia. Based on published data, we used a national average of $46 400 cost of care per postoperative pneumonia case to estimate potential cost savings in our analyses. The validity of cost estimates in this study to individual hospitals must be considered because geographic variation in cost of care is well established. One of the challenges of sustaining the pneumonia prevention program is turnover of nursing staff and trainees (residents in particular) on the surgical ward. Although the impact of such turnover on the degree of success of the program is unclear, we were able to record sustained decrease in pneumonia rates despite this reality. Indeed, to alleviate the potential negative impact of staff turnover on our program, 2 policies were enacted at the program’s inception and have been maintained. First, nursing leadership mandated that prevention interventions were part of nursing staff standard work and were included in a nurse’s daily assessments. Program adherence is charted in each shift’s nursing progress note. Second, an electronic order set was imbedded in all admitting surgical services order sets. Medically relevant choices such as activity level and head-of-bed elevation default to “ambulate” and “bed elevation of 30°” but the physician can change those choices. Options such as mouth care, spirometry, and coughing/deep breathing instructions autoload without physician input.
Another potential limitation of this study is the lack of adjustment for case mix. In 2008, the VA mandated that VA facilities that perform surgical procedures would be classified as complex, intermediate, or basic facilities based on case mix and facility infrastructure.10 This policy later led to mandated transfer of specific complex cases out of non–complex-designated facilities to complex facilities. The VA Palo Alto is a long-standing tertiary hospital and a designated complex hospital, thereby defining it as the regional referral center. Therefore, the VA Palo Alto case mix is similar to that of tertiary care medical centers except for the inclusion of typical level I trauma interventions (data for the latter also are not included in ACS-NSQIP). We believe that the overall case mix at VA Palo Alto is similar to that of ACS-NSQIP, since ACS-NSQIP contains a broad inclusion of various hospital settings. In addition, every hospital included in the VA-SQIP receives their quarterly and annual distribution of case complexity (complex, intermediate, or basic); the case distribution at VA Palo Alto did not substantially change over the study period. Another potential limitation is the primarily male population at the VA. To account for this, the cohort selected for comparison was limited to males within ACS-NSQIP. Moreover, recent analyses of pneumonia risk based on ACS-NSQIP data11 found that sex was not a significant factor in overall pneumonia risk; therefore, we do not believe that limiting the comparison cohort to male patients significantly affected our results. Finally, because this intervention was performed at a single medical center, results are potentially not generalizable to other hospitals.
Despite the limitations listed earlier, our study supports the concept that successful and sustained reduction of pneumonia among postoperative patients requires multiple performance measures and unrelenting standardized quality improvement efforts.
Accepted for Publication: March 28, 2014.
Corresponding Author: Sherry M. Wren, MD, Stanford University School of Medicine, Veterans Administration Palo Alto Health Care System, 3801 Miranda Ave, G112, Palo Alto, CA 94304 (firstname.lastname@example.org).
Published Online: July 23, 2014. doi:10.1001/jamasurg.2014.1216.
Author Contributions: Drs Wren and Kazaure had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Kazaure, Yoon, Wren.
Acquisition, analysis, or interpretation of data: Kazaure, Martin, Wren.
Drafting of the manuscript: Kazaure, Wren.
Critical revision of the manuscript for important intellectual content: Martin, Yoon, Wren.
Statistical analysis: Kazaure, Wren.
Administrative, technical, or material support: Martin, Wren.
Study supervision: Yoon, Wren.
Conflict of Interest Disclosures: None reported.
Disclaimer: The ACS-NSQIP and the hospitals participating in the ACS-NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.
Previous Presentation: Data in this study were presented at the Pacific Coast Surgical Association meeting; February 16, 2014; Dana Point, California.