Case Study High-Performing Health Care Organization • March 2009 October 2009 Holland Hospital: Improving Pneumonia Care by Hardwiring Process Enhancements B y A imee L ashbrook , J.D., M.H.S.A. H ealth M anagement A ssociates The mission of The Commonwealth Vital Signs Fund is to promote a high performance Location: Holland, Mich. health care system. The Fund carries out this mandate by supporting Type: Private, nonprofit hospital independent research on health care Beds: 209 issues and making grants to improve health care practice and policy. Support Distinction: Top 3 percent in composite of seven pneumonia process-of-care measures, among more for this research was provided by than 2,800 hospitals (more than half of U.S. acute-care hospitals) eligible for the analysis. The Commonwealth Fund. The views Timeframe: July 2007 through June 2008. See Appendix for full methodology. presented here are those of the author This case study describes the strategies and factors that appear to contribute to high performance on and not necessarily those of The pneumonia process-of-care measures at Holland Hospital. It is based on information obtained from Commonwealth Fund or its directors, interviews with key hospital personnel, publicly available information, and materials provided by the officers, or staff. hospital during June through September 2009.      Summary For more information about this study, Holland Hospital has significantly improved its performance on the pneumonia please contact: process-of-care, or “core” measures, over the last five years. The core measures, Aimee Lashbrook, J.D., M.H.S.A. Health Management Associates developed by the Hospital Quality Alliance, relate to provision of recommended alashbrook@healthmanagement.com treatment in four clinical areas: heart attack, heart failure, pneumonia, and surgi- cal care. Holland Hospital performs in at least the top 20 percent in all four areas, and particularly well in pneumonia and surgical care. When the Centers for Medicare and Medicaid Services (CMS) adopted the pneumonia care core measures in 2004, Holland Hospital was achieving only 50 To download this publication and learn about others as they become to 60 percent compliance on some of them. Since then, it has become one of the available, visit us online at top performers in the country in terms of delivering recommended pneumonia www.commonwealthfund.org and care. To improve performance, Holland Hospital made process improvements register to receive Fund e-Alerts. and hardwired them into its electronic medical record system. Strong manage- Commonwealth Fund pub. 1327 Vol. 30 ment support and a core measures leadership team dedicated to providing root- 2T he  C ommonwealth F und cause analysis, oversight, and direction also played a team identified ways to automate processes and enable significant role. performance reporting, hospital leaders provided the support needed to make such changes. These advance- Organization ments have improved performance on all of the core Holland Hospital, in Holland, Michigan, has 209 measures, not just the pneumonia set. acute-care beds; it is not part of a larger health care system. Holland Hospital’s medical staff includes over Core Measures Teams 300 physicians, representing 34 medical specialties. A core measures leadership team oversees the hospi- Holland Hospital serves more than 7,500 inpa- tal’s performance on these measures. Because the team tients each year, with 43,100 annual emergency room includes physicians as well as clinical directors and visits, 367,400 annual outpatient visits, 28,600 urgent other leaders, it helps secure physicians’ buy-in for care visits, and 9,200 surgeries. In 2007, Holland new initiatives. became the first hospital in Western Michigan to earn The team reviews “opportunities for improve- Magnet designation, an honor recognizing excellence ment” (OFIs), including cases involving noncompli- in nursing and patient care. It has received other hon- ance with core measures, on a monthly basis. Instead ors and designations, including the Governor’s Award of focusing on the clinicians involved in a noncompli- of Excellence for Improving Care in the Hospital ant case, the team focuses on the system factors that Setting, the Total Benchmark Solution Best Acute Care may have contributed to the error. According to Hospitals Award, and the National Kidney Foundation Schwartz, “the hospital’s patient safety culture means of Michigan Innovations in Health Care Award. For being blame-free. Unless the case is egregious, we the last four years, Holland Hospital also has been assume mistakes occurred because the established care listed as one of the “100 Top Hospitals” by Thomson process failed our staff and/or physicians.” Reuters (formally Solucient). In addition, the hospital created teams focused on core measures related to cardiology, surgery, and Hospital-Wide Strategies respiratory disease (discussed below). Beginning in 2005, when CMS began publicly report- ing hospital performance data and Holland Hospital Concurrent Review hired a new director of quality and risk, the hospital All patients are assessed for possible inclusion in the has been paying close attention to the core measures— core measure population. The assessment is built into working them into its information systems, education, the hospital’s electronic medical record (EMR), a sys- and employee financial incentives. tem known as Quadramed CPR. When patients exhibit Using the MIDAS system, Holland Hospital symptoms of a core measure diagnosis, their nurse will sets internal benchmarks against which to measure its flag them as potential core measure patients. This performance.1 Each year, the bar is set higher, with tar- automatically notifies analysts in the quality depart- gets moving from the 75th to the 90th percentile. ment, who initiate a concurrent review. One analyst is Rob Schwartz, M.B.A., M.H.A., M.S./M.I.T. dedicated to reviewing the pneumonia care core (master of science in the management of information measures. technology), director of quality and risk, credits Clinical managers and directors also receive Holland’s administrators with providing the support daily status reports about the patients in the core mea- needed to raise performance levels and create a culture sure population. Keeping multiple pairs of eyes on of quality improvement. Schwartz drew on his back- these records means that noncompliant cases can be ground in information technology to strengthen the flagged and most issues can be addressed before hospital’s computer systems. As he and his patients leave the hospital. H olland H ospital : I mproving P neumonia C are by H ardwiring P rocess E nhancements 3 Financial Incentives for All Staff embedded in the hospital’s patient safety culture. All full-time staff members, from janitors to emer- Rather than blame an individual for a case that falls gency room clinicians, have a stake in the hospital’s out of compliance, leaders look for failures in the performance on the core measures—demonstrating established care processes and seek potential system that all staff have a role to play in quality solutions. For example, when Holland’s quality staff improvement. found that nurses were not consistently screening for Holland’s goal-sharing program targets core and administering the pneumonia vaccine, they real- measure performance as well as patient satisfaction ized that a system factor was contributing to the fail- scores. Each year in which the hospital achieves a ures. “The electronic nursing record was allowing margin that exceeds budgeted projections, a bonus nurses to skip the vaccine screening questions,” pool is established. If certain performance goals are Schwartz says. “Once we reprogrammed the computer met, up to $500,000 will be distributed from the bonus system to require an answer to the vaccine screening pool. This can amount to an annual bonus of approxi- questions before the nurse could proceed, our vaccina- mately $300 to $400 for each staff member. The tion administration scores improved tremendously.” amount of funds distributed varies, based on the hospi- Despite the focus on such system-based solu- tal’s performance on targeted quality indicators. tions, physicians and other hospital staff are held However, if the hospital fails to meet a threshold per- accountable when necessary. Some cases require one- formance level, no funds are distributed. This year, the on-one meetings with noncompliant staff to provide threshold performance level for the core measures was education and ensure follow-up steps are taken. Before set at 90 percent compliance and the “stretch” goal, the meeting, quality department staff will make certain which triggers a greater distribution of funds, was set that the case was in fact noncompliant and not merely at 96 percent. The targets have increased since the first a mistake in documentation or some other system- year of the program, when the stretch performance related error. level was set at 90 percent. Individual physician performance is monitored When bonuses are available, they are incorpo- and compiled in a Physician Feedback Report. The rated into annual staff performance reviews. Staff quality department compiles this report and shares it members who do not meet a certain performance level with individual physicians. Any outliers are forwarded or are on a corrective action plan are not eligible. to the Peer Review Committee Chair and the Vice presidents and other hospital executives are also Credentialing Committee Chair for review. The report not eligible. includes numerous indicators, including performance All patient cases, not just Medicare cases, are in the core measures, patient safety, and citizenship assessed for compliance with the core measures. (e.g., participation in hospital committees, presenta- Because hospital leaders believe that providing all rec- tions, grand round lectures, and similar activities.). ommended care at all times is the true test of quality, performance on “all-or-none” care bundles is also Pneumonia Care Improvement tracked. Under this measure, only cases in which the Strategies care delivered meets each applicable core measure are When the pneumonia core measures were introduced counted as compliant. in 2004, Holland Hospital created a respiratory disease core measure team to focus on care processes related A Focus on System Factors to them. The team, which included pharmacists, physi- There is an underlying philosophy at Holland Hospital cians, care managers, quality department staff, nurses, that core measure performance is strongly linked to and hospital leaders (such as the pharmacy director system factors and solutions. This philosophy is and emergency room director), developed many of the 4T he  C ommonwealth F und improvement strategies outlined below. A physician patient has had the appropriate vaccinations. The leader naturally emerged, helping to guide the team’s record will not move forward until the answers are efforts and keep energy levels high. obtained. This feature can be turned on and off accord- ing to the season. The EMR also keeps smoking cessa- Hardwiring Quality Improvement tion reminders visible until a staff member indicates Holland Hospital made many changes to its care pro- that the required counseling has been provided. cesses to improve performance in the pneumonia care core measures. The respiratory disease core measure Shifting Responsibilities team developed a preprinted order set to guide physi- When the pneumonia core measures were introduced, cians in the provision of care, especially with the use Holland Hospital staff struggled with one measure in of appropriate antibiotics. In developing the order set, particular: blood culture prior to initial antibiotic the team solicited feedback from internists and emer- administration. There were frequent delays in taking gency room physicians—the physicians whom they blood cultures, in part because phlebotomists had to be considered to be its primary users. called to take the blood and in part because staff often Development of the order set was facilitated by mistakenly assumed that a blood culture had been a physician champion, although even then it was diffi- taken if an IV infusion had been started. To avoid cult to achieve consensus. Holland’s physicians main- these problems, responsibility for taking blood cultures tained that, because patients could easily fall in and was transferred from phlebotomists to emergency out of the criteria during a hospital stay, physicians room nurses. Because the nurses are also responsible should not be bound to a preset clinical pathway. for administering antibiotics, they can control the Eventually, the physicians agreed to try the new sys- order in which the two interventions occur. When nec- tem. Patients who exhibited symptoms indicating a essary, nurses can reach a phlebotomist devoted to the strong risk of pneumonia were started on an evidence- emergency room over the hospital’s instant communi- based clinical pathway. Joe Bonello, R.N., director of cations system. emergency services, found that “pharmacists and In addition, the hospital is developing a system nurses, empowered by the order set, were more likely that will place a “hold” in the pharmacy on antibiotic to challenge orders that were outside of the recommen- orders for pneumonia patients until blood cultures are dations and suggest items that appeared on the order documented in the EMR. set. This led to a change in physician ordering Similarly, hospital leaders discovered that patterns.” patients admitted to medical units from the emergency Once finalized, the pneumonia care order sets department experienced delays in antibiotic adminis- were hardwired into the hospital’s EMR. The system tration. Therefore, they decided to have the initial requires physicians to document any departure made course of antibiotics administered and documented in from the order set and explain their reasoning. the emergency room, making it much easier to meet According to Bonello, “a good order set minimizes the the core measure standard requiring antibiotic adminis- unwarranted variation in decision-making through tration within six hours of arrival. Charge nurses in the standardization and reduces the potential for emergency room perform real-time chart audits to noncompliance.” ensure the antibiotic has been administered and Reminders and other clinical pathways are built required documentation has been captured before the into the hospital’s EMR. For example, a vaccination patient leaves the emergency room. assessment is included in the nurse’s assessment Finally, in cases where patients meet certain cri- screen. If a patient meets certain age criteria, the teria, nurses have been given the power to administer assessment screen will prompt the nurse to ask if the vaccinations without a physician’s order. As noted H olland H ospital : I mproving P neumonia C are by H ardwiring P rocess E nhancements 5 Figure 1. Holland Hospital Scores on Pneumonia Care Core Measures Compared with State and National Averages National Michigan Pneumonia Care Improvement Indicator Average Average Holland Hospital Percent of pneumonia patients given oxygenation assessment 99% 100% 100% of 265 patients Percent of pneumonia patients assessed and given 83% 85% 100% of 240 patients pneumococcal vaccination Percent of pneumonia patients whose initial emergency room 90% 92% 99% of 233 patients blood culture was performed prior to the administration of the first hospital dose of antibiotics Percent of pneumonia patients given smoking cessation 88% 90% 100% of 62 patients advice/counseling Percent of pneumonia patients given initial antibiotic(s) within 93% 95% 100% of 173 patients six hours after arrival Percent of pneumonia patients given the most appropriate 87% 91% 100% of 121 patients initial antibiotic(s) Percent of pneumonia patients assessed and given influenza 79% 84% 99% of 144 patients vaccination Source: www.hospitalcompare.hhs.gov. Data are from July 2007 through June 2008. above, a vaccination assessment is hardwired into the Results nursing assessment screen of the EMR system. Making Holland Hospital outperforms most U.S. hospitals on vaccination administration part of the EMR and waiv- all of the pneumonia care core measures. Figure 1 dis- ing the need for physician approval helps ensure com- plays the most recent year of data, while Figure 2 pliance with recommended care. shows the trends over time for the all-or-none pneu- monia care bundle. According to Schwartz, “success Education, Education, Education breeds success”; each year it has become easier to Holland Hospital leaders emphasize the importance of make further improvements. For example, the hospital education in improving performance. In conjunction experiences so few cases that fall out of compliance with hardwiring new processes, the hospital devoted that members of the core measure leadership team, substantial resources to educating staff, particularly including vice presidents and clinical leaders, can emergency room staff, about the core measures. Topics review each of them and develop strategies for solving included the importance of performing a blood culture underlying problems. prior to the administration of antibiotics and the need As shown in Figure 2, Holland Hospital experi- to start antibiotics in the emergency room to ensure enced a dip in its performance in the all-or-none pneu- they are given within six hours of arrival. Evidence- monia care bundle in 2004 and 2005, with solid based literature was used to convince physicians of improvement thereafter. This was related, in part, to the effectiveness of a new practice. In addition, the introduction of new pneumonia care core measures social workers and nurses were trained to identify within that timeframe (antibiotics within four hours of smokers who, for whatever reason, do not indicate arrival, appropriate antibiotic selection, and the influ- during registration that they smoke. This helps enza vaccine). According to Schwartz, the new mea- ensure that all patients who need it are given sures caught the hospital “off guard.” Now, when a smoking cessation counseling. new measure is announced, usually six to 12 months 6T he  C ommonwealth F und Figure 2. Holland Hospital Scores on "All-of-None" Pneumonia Care Bundle, 2003–2008 Percent 96 99 100 80 75 65 66 59 60 40 20 0 2003 2004 2005 2006 2007 2008 Note: All-or-none bundles include all seven pneumonia care core measures. Source: Holland Hospital, 2009. in advance of data collection, the hospital immediately • Interdisciplinary workgroups can focus on begins to track their performance, giving clinicians an opportunities to improve care systems—rather opportunity to make improvements before reporting than blame individuals for problems. data to CMS. The hospital strives to be proactive in • A goal-sharing program linked to core mea- improving performance, paying attention to the activi- sure performance helps motivate employees ties of the National Quality Forum and the Agency for and establishes a culture focused on quality Healthcare Research and Quality. It also reviews the improvement. proposed and final Inpatient Prospective Payment System rules issued by CMS to identify clinical areas, Holland Hospital is now consistently perform- such as stroke care, that the agency might focus on in ing at a high level in the pneumonia care core mea- future public reporting efforts. sures. Leaders no longer look for “silver bullets” to solve problems, such as shifting the responsibility for Challenges and Lessons Learned taking blood cultures or reprogramming its electronic Hospitals looking to achieve high performance in the nursing record. Instead, they carefully examine the few pneumonia care core measures might take the follow- cases that fall out of compliance and find opportunities ing lessons from Holland’s experience: to eliminate the causes. Schwartz refers to the Swiss cheese model of quality improvement in explaining the • Support from upper management is key. need to “plug holes.”2 “Our OFIs are all latent prob- lems that manifest themselves when certain holes line • Flagging core measure patients and perform- up,” he says. “We continually expect to have more ing concurrent review greatly reduces the like- OFIs, usually caused by a unique set of circumstances, lihood of a case falling out of compliance. but we continue to try to engineer these circumstances • A core measures leadership team can be used out of our processes. Upon analysis we often find that to monitor performance and maintain the the circumstances that produced the OFI are unlikely momentum for performance improvement efforts. to happen again, but we still try to prevent the reoccurrence.” H olland H ospital : I mproving P neumonia C are by H ardwiring P rocess E nhancements 7 Holland Hospital staff note the challenge of For More Information keeping energy levels high. By having monthly meet- For further information, contact Rob Schwartz, M.B.A., ings of the core measures leadership team, the organi- M.H.A., M.S./M.I.T., director of quality and risk, at zation maintains its focus on quality improvement. Rschwartz@hollandhospital.org. Holland Hospital faces some challenges that are beyond their control. For example, the recommended antibiotics are at times difficult to obtain from the manufacturer, an issue that could result in cases unnec- essarily falling out of compliance with the pneumonia care core measures. Also, Holland Hospital uses inter- nal data to assess the community’s pneumococcal resistance to antibiotics, which at times results in the need for an antibiotic regimen that contradicts the core measure recommendations. In these cases, Holland Hospital has reached out to their state Medicare Quality Improvement Organization and asked it to raise their concerns with CMS. 8T he  C ommonwealth F und N otes 1 The MIDAS system is an integrated medical infor- mation management system for hospitals that en- ables comparative data analysis and clinical bench- marking using a large concurrent database with over 3,000 clinical metrics. See http://www.midasplus. com/index.asp. 2 James Reason developed the model to illustrate how smaller systems failures combine to create an error. In the model, individual slices of cheese represent protections against error. The holes in the cheese, which vary in size and position, represent individual mistakes. When the holes align, an error occurs, such as a case falling out of compliance. H olland H ospital : I mproving P neumonia C are by H ardwiring P rocess E nhancements 9 Appendix. Selection Methodology Selection of high-performing hospitals in process-of-care measures for this series of case studies is based on data submitted by hospitals to the Centers for Medicare and Medicaid Services. We use seven measures that are publicly available on the U.S. Department of Health and Human Services’ Hospital Compare Web site, (www.hospitalcompare.hhs.gov). The measures, developed by the Hospital Quality Alliance, relate to practices in pneumonia care. Pneumonia Care Process-of-Care Measures 1. Percent of pneumonia patients given oxygenation assessment 2. Percent of pneumonia patients assessed and given pneumococcal vaccination 3. Percent of pneumonia patients whose initial emergency room blood culture was performed prior to the administration of the first hospital dose of antibiotics 4. Percent of pneumonia patients given smoking cessation advice/counseling 5. Percent of pneumonia patients given initial antibiotic(s) within six hours after arrival 6. Percent of pneumonia patients given the most appropriate initial antibiotic(s) 7. Percent of pneumonia patients assessed and given influenza vaccination The analysis uses all-payer data from July 2007 through June 2008. To be included, a hospital must have sub- mitted data for all seven measures (even if data submitted were based on zero cases), with a minimum of 30 cases for at least one measure, over four quarters. The top 3 percent among 2,887 hospitals eligible for the analysis and with 50 or more beds were considered high performers. In calculating a composite score, no explicit weighting was incorporated, but higher-occurring cases give weight to that measure in the average. Since these are process measures (versus outcome measures), no risk adjust- ment was applied. Exclusion criteria and other specifications are available at http://www.qualitynet.org/dcs/ContentS erver?cid=1141662756099&pagename=QnetPublic%2FPage%2FQnetTier2&c=Page). While a high score on a composite of surgical care improvement process-of-care measures was the primary criterion for selection in this series, the hospitals also had to meet the following criteria: at least 50 beds, not a gov- ernment-owned hospital, not a specialty hospital, ranked within the top half of hospitals in the U.S. in the percent- age of patients who gave a rating of 9 or 10 out of 10 when asked how they rate the hospital overall (measured by Hospital Consumer Assessment of Healthcare Providers and Systems, HCAHPS), full accreditation by the Joint Commission, not an outlier in heart attack and/or heart failure mortality, no major recent violations or sanctions, and geographic diversity. A bout the A uthors Aimee Lashbrook, J.D., M.H.S.A., is a senior consultant in Health Management Associates’ Lansing, Mich., office. Ms. Lashbrook has six years of experience working in the health care industry with hospitals, managed care organizations, and state Medicaid programs. She provides ongoing technical assistance to state Medicaid pro- grams, and has played a key role in the development and implementation of new programs and initiatives. Since joining HMA in 2006, she has conducted research on a variety of health care topics. Aimee earned a juris doctor degree at Loyola University Chicago School of Law and a master of health services administration degree at the University of Michigan. A cknowledgments We wish to thank Rob Schwartz, M.B.A., M.H.A., M.S./M.I.T., director of quality and risk; Joe Bonello, R.N., director of emergency services; Bill Brackenridge, Pharm.D., M.A., director of pharmacy; and Gary Harrison, R.N., performance improvement coordinator, for generously sharing their time, knowledge, and materials with us. Editorial support was provided by Martha Hostetter. This study was based on publicly available information and self-reported data provided by the case study institution(s). The Commonwealth Fund is not an accreditor of health care organizations or systems, and the inclusion of an institution in the Fund’s case studies series is not an endorsement by the Fund for receipt of health care from the institution. The aim of Commonwealth Fund–sponsored case studies of this type is to identify institutions that have achieved results indicating high performance in a particular area of interest, have undertaken innovations designed to reach higher performance, or exemplify attributes that can foster high performance. The studies are intended to enable other institutions to draw lessons from the studied institutions’ experience that will be helpful in their own efforts to become high performers. It is important to note, however, that even the best-performing organizations may fall short in some areas; doing well in one dimension of quality does not necessarily mean that the same level of quality will be achieved in other dimensions. Similarly, performance may vary from one year to the next. Thus, it is critical to adopt systematic approaches for improving quality and preventing harm to patients and staff.