Mending Broken Hearts
(Feb. 12, 2009) Brent Hawkins sat down to breakfast in a Minnesota Marriott last summer. A Eureka resident for the past eight years, Hawkins heads a division of the agricultural giant Cargill. He travels frequently. Minnesota is his territory; USA Today his breakfast companion.
It was Aug. 21 — the official unveiling of the government’s best estimates of risk-adjusted death rates for heart attack, heart failure and pneumonia patients in every U.S. hospital. With a wife on the St. Joseph Hospital Foundation board and St. Joseph physicians among his friends, Hawkins is an informed health consumer, someone who would be interested in a story about the “extreme differences between the worst and the best” hospitals. And he was.
“I was shocked to find our hospital among the worst in the country,” Hawkins said. “Everybody kind of expected us to be average.”
The paper reported that the government’s estimated 30-day heart failure death rate for St. Joseph Hospital in Eureka is 15.4 per 100 Medicare patients served from July 1, 2006 through June 30, 2007. The national rate is 11.2. (The figures are adjusted for the age, income and general sickness level of each hospital’s patient pool.) Despite a statistically conservative approach that labeled less than 1 percent of all hospitals as better or worse in a particular category, St. Joseph Hospital is one of 35, nationwide, to find itself on the “worse” list for heart failure.
“Four more deaths is a big difference,” Donald Berwick said last month from his office in Cambridge, Mass. Berwick is president and chief executive officer of the Institute for Healthcare Improvement, a nonprofit center working with hospitals to improve the quality of care and eliminate errors.
“Most people who die are destined to die because they have things we can’t change the trajectory of,” Berwick said. “Health care doesn’t affect mortality that much, so you are looking at a huge difference in the impact of health care. It’s very unlikely the difference is random. Something is probably there.”
Hawkins would like to know what.
“I haven’t even heard them acknowledge this ratings issue, and I’ve heard nothing about what they are doing to fix it.”
Publicizing bad news in the midst of a $120 million fund-raising campaign for a hospital expansion is hard to do, even though Berwick said “going public will earn you credit.”
It’s harder yet when hospital personnel harbor concerns with the study: the small sample size of 50 patients and eight deaths, the controversy over the use of risk-adjusted mortality data and the improper inclusion of subjects as a result of the hospital’s past failures in data collection and coding.
“I’m not convinced 15.4 is our number, but that’s a hard thing to say when you don’t have a handle on the numbers to back up your assumptions,” said Dr. David Ploss, cardiologist and former chief of staff at St. Joseph Hospital in December.
Even though it doubts the findings, the hospital leadership investigated them. It contends that the numbers will be improved in the next go-round because St. Joseph is a hospital in transition, inside and out.
Heart failure is a chronic disease and the most common cause of hospitalization among those on Medicare, which is why, in the study that Hawkins discovered, Medicare made it one of its core focus areas for quality improvement. The disease refers to the heart’s inability to pump enough blood and oxygen to the body’s other organs. Shortness of breath or swelling in the legs, ankle and abdomen are common symptoms.
Five million people in the U.S. cope with heart failure. The Centers for Disease Control and Prevention reports 550,000 people are diagnosed each year, more than enough to replace the 287,000 people who die.
“You can’t fix heart failure. It doesn’t go away,” Ploss said. “It’s manageable, not fixable. What you try to do is make sure patients can live safely and not bounce back and forth between home and the hospital.”
The American Heart Association and National Heart, Lung and Blood Institute report 22 percent of men and 46 percent of women will develop heart failure within six years of having a heart attack and seven in 10 people with heart failure had high blood pressure before being diagnosed. Not every death is a failure or quality problem. One study puts the average life expectancy after a diagnosis of heart failure at a little over three years for men and five years for women.
“Many [heart failure patients] tend to be frequent fliers, and they jack up the cost of care tremendously,” said Dr. Prediman K. Shah, director of cardiology at Cedars-Sinai Medical Center in Los Angeles, last month. “For a hospital, it’s a fiscally and programmatically important group.”
It’s not an easy condition to treat, but some hospitals clearly do better than others. Cedars-Sinai is one of 19 hospitals in the country tabbed by U.S. News & World Report as the Best of the Best. With a risk-adjusted mortality rate of 7.1, the Medicare study identified it among the 41 hospitals in the U.S. with “better” heart failure rates.
Shah said Cedars-Sinai prospectively tracks from admission to discharge all patients admitted with a diagnosis of heart failure to make sure an appropriate series of tests and algorithms mandating certain treatments and medications are used to positive effect.
“If something hasn’t been done, it’s a red flag and there are electronic reminders,” he said.
Shah is one of 16 hospital-based cardiologists working full-time at Cedars-Sinai, a research and teaching hospital that caters to Hollywood and the upper crust. He said another 90 cardiologists practice privately. Access to care probably wasn’t the problem for the 382 patients from whom the Cedars-Sinai heart failure data came.
“By and large, we have an educated clientele because of our location,” Shah said. “Patients tend to be pretty responsible and quite compliant after discharge.”
Other pluses: “educated, seasoned internists” exposed to numerous educational activities so they become sophisticated at managing heart failure. “If you want good outcomes, you can’t rely on cardiologists alone,” Shah said. “A lot of the time, internists, nephrologists and family practitioners are the ones taking care of heart failure patients.”
In contrast, Humboldt County’s population is seldom described as affluent or educated and has been repeatedly warned by the medical society of a physician shortage in primary and specialty care. There are no hospital-based cardiologists at St. Joseph and three cardiologists in the county to split time between private practice and the hospital.
Yale cardiologist Dr. Harlan Krumholz said in October that it is “unrealistic” to expect a rural hospital to duplicate a resource-rich teaching and research institution, but that lessons can be found in the search to explain variation in outcomes.
“There is not a perfect hospital right now,” Krumholz said. “It’s implausible to think that based on the state of care, there isn’t room to improve.
“Serious people when they look hard at care can find ways to make treatments more effective: to give the right doses at the right times, avoid wrong medications, wash hands, provide orders in timely ways and make sure communication between providers and patients is done well.
“The question is: Are they digging in?”
The St. Joseph Hospital community has been digging in for the past three years.
The heart failure data — now 20 to 31 months old — comes from the first fiscal year after Joseph Mark took over as chief executive officer in February 2006.
Awash in red ink, Mark came on board to dig out of a financial hole that had SJH on pace to lose $14 million in one year. The hospital posted a 4 percent net profit last fiscal year.
“We weren’t working on operational efficiency for a few years,” Ploss said of what he termed the “headless” period. “The focus was on the money needed to stay open.”
Spring layoffs followed, and Mark began implementing efficiencies to approach benchmarks set by other hospitals.
The bottom line wasn’t the only challenge.
The Joint Accrediting Commission for Healthcare Organizations gave St. Joseph Hospital only a conditional accreditation after a survey of the SJH laboratory in April 2006. Former JACHO media relations director Charlene Hill said in fall 2006 the agency conducted 1,450 voluntary hospital audits annually, denying accreditation to 1 percent and conditionally accrediting 2 percent.
A hospital committee spent a year making sure 2008 accreditation went without a hitch, but the hospital spent much of 2006 responding to JACHO demands to improve in 10 areas dealing with workload limits, tissue testing, equipment maintenance, recordkeeping, comparable treatment and employee competency. (In addition, the JACHO report required an improvement in the laboratory environment and effectiveness of communication among caregivers.)
Mark’s administrative challenges included a narrowly averted strike with the nurses and a less-publicized standoff with his on-call general surgery doctors. The general surgeons refused to work at St. Joseph for a little over four months early in 2007, leaving Drs. Ellen Mahoney and Luther Cobb to cover calls for a tertiary care hospital that serves all or parts of four counties and southwestern Oregon. Tory Starr, director of performance improvement and quality management for St. Joseph, admits infection control and discharge instructions were also problematic.
Discharge instructions — follow-up guidelines given to patients after they leave the hospital — are particularly important in the case of heart failure. Medicare desires discharge instructions to help patients manage heart failure with information about activity levels, diet, medications, daily weight monitoring, follow-up appointments and actions to take if symptoms get worse.
“A lot of it is patient education,” said St. Joseph Hospital intensive care nurse Kathryn Donahue. “Some wait too long and have to be intubated. Why didn’t they come in days before when their feet were swollen? Not having insurance may have something to do with it.”
Prior to Mark’s arrival, Medicare reported St. Joseph Hospital gave discharge instructions to 23 percent if its heart failure patients compared to 47 percent statewide, 52 percent nationally and 88 percent of the top hospitals. For the nine months ending in March 2008, St. Joseph boosted that to 52 percent.
Starr expects that number to climb as the hospital now uses standardized order sets with all the necessary elements to eliminate confusion for nurses about which discharge instructions to use.
“But can (patients) read and understand or are you just trying to get boxes checked and paperwork done?” asked Donahue. “Medical stuff is complicated and nurses are the translators. You have to go in, sit down and go at their speed. That’s very time consuming, a drain that’s not allowed for or staffed, so patients lose out.”
In 2007, Starr began the Care Transitions Program to fill the gap for all patients who move from acute care or outpatient settings to home. In partnership with Humboldt State University’s nursing program, the California HealthCare Foundation, and the Robert Wood Johnson Foundation, St. Joseph’s outreach involved 100 discharged patients by December. Fourth-year student nurses serve as coaches and case managers. They meet with potential clients prior to discharge then make home visits and phone calls to build a personal health record that can follow the patient to her permanent medical home, reconcile the medications list, teach about “red flag” warning signals and prep patients for follow-up discussions with their primary care providers.
“We haven’t run the stats yet, but we’ve seen some good results in cutting down multiple ER visits and managing medications,” said Sharon Hunter, the nurse St. Joseph hired in July to devote 80 percent of her time to the program.
Starr reported Care Transitions follow-up discovered one or more medication errors in nearly 75 percent of the client visits and 6 percent had five to six medication errors. “Medication errors are a well-documented, preventable cause of readmission,” Starr said.
“I had a list of medications, discharge instructions and so many questions,” Care Transitions participant Dorelee Heisler said. “(The program) gave me peace of mind. If you feel terrible from the drugs and wonder if you should stop taking them or need to find a dietitian like I did, there’s someone to call, someone to help me know how to ask questions when I saw my doctor.”
Arcata physician Dr. Alan Glaseroff, medical director of the Humboldt-Del Norte Independent Practice Association, said he’s seen improvement lately as hospital discharge units and physicians he knows attempt to keep primary care physicians in the loop with lab results or medications lists.
“It was not uncommon for one of my patients to show up at St. Joe or Mad River and I’d never know about it,” he said of the not-too-distant past. “I’d frequently have patients leave the hospital and fill new prescriptions without talking to me.”
There’s little doubt that improved communication between hospital, physician and patient saves lives. Eight months ago, Glendale Memorial Hospital in Southern California recruited registered nurse Susan Thompson from the emergency room at USC Medical Center to run its chronic disease management program for heart failure and heart attack.
The 334-bed facility is one of the state’s top performers with a 7.7 mortality estimate for heart failure, but it will soon add to her duties a Web-based, telephone-monitoring program to track discharged patients she’s recruited.
The program is free, grant-funded and allows patients to call in and answer a series of questions. Changes pop up on a daily report and Thompson calls to follow-up.
“Elderly folks won’t shell out $3,000 bucks for some in-home system,” she said. “They will talk to a nurse on the phone.”
The goal: 100 clients this year and 200 in 2010.
Thompson also evaluates each new hospital admission and looks at the notes through the entire admission to capture anyone who might fall into a heart failure diagnosis.
“She’s a clinical specialist who works right alongside nurses as a resource person, mentor, teacher and patient liaison,” Donahue said. “We don’t have that here and we haven’t had that in years. It’s one of the big failings.”
Starr said the hospital talks about the Medicare core patients every day at bed control meetings and in the individual nursing units, and tracks their interventions.
But Ploss said that there’s still a gap in the tracking. “Cardiologists see only a minority of heart failure patients admitted to the hospital,” he said. “Evaluating readmission is key and we are not doing that in any way that is reported back to the physicians taking care of the patients.” St. Joseph Hospital has upgraded its electronic medical records, but has no electronic reminder system, Ploss said.
While the hospital tackles the heart failure challenges, the community can look to the intensive care unit and cardiac surgery for hope that St. Joseph is up to the task.
Though the hospital may lag behind in treatment of heart failure and some other areas of medicine, it has reason to be very proud of the gains it has made in others. The question is whether it can apply the lessons of its most successful programs across the entire spectrum of care.
“This hospital, for the first time, is in a renaissance,” said Dr. Joachim Postel, the cardiac surgeon who arrived from Miami to lead the SJH Heart Institute in 2000.
Glaseroff agrees. He’s one of the drivers in the Robert Wood Johnson-funded effort to improve the care of chronic disease in Humboldt County and a national leader in setting performance measures to improve the quality of care.
“They are doing a lot of work to improve the last two years,” he said. “I give Tory (Starr) and the hospital a lot of credit.”
The key: introducing standardized processes developed from evidence-based guidelines. “Once there’s a protocol, a system in place, it creates a repetitive model,” Postel said. “If you have complications — problems — then it’s much easier to look back and detect or manage them. If something didn’t work, it can be changed.”
Writing out a protocol challenges a physician’s thinking and knowledge, Postel said. “You have to look in the literature to create guidelines,” he said. “It’s no longer about everyone doing something in their own way.”
When he arrived, there was no protocol for diabetics, despite the fact that high blood sugar correlates with infections, he said. He’s built a system of algorithms and procedures that brought statewide recognition for good outcomes to a cardiac surgery program that got off to a rocky start in the early 1990s, when a much higher than expected mortality rate prompted an investigation and a temporary shutdown.
Since 2000, statistics reported to the Society of Thoracic Surgeons and the California Coronary Bypass Outcome Report credit Postel with 785 cardiac surgeries and an unadjusted mortality rate of 1.5 percent. He hasn’t lost a single patient in 122 valve and aorta surgeries, despite the 3.5 to 4 percent mortality for the simplest aortic valve replacement and 8 to 12 percent for the most difficult.
“That’s where you separate the men from the boys,” Postel said with a precise German accent and a hint of pride.
“He’s an excellent surgeon who follows his patients like a mother hen before they come in until the day they leave and after,” Donahue said. “He couldn’t do that anywhere else, where the surgeons are relieved after the surgery. But we don’t have a high volume of patients, maybe one or two a week, so that’s made that possible.”
Postel told the hospital board recently that the cardiac team’s work and refined protocols have dropped the percentage of complications in coronary bypass surgery from 8.2 percent in 2006 to 4.8 percent in 2007 to 0 percent last year. In comparison, he said statewide and national numbers for complications exceed 27 percent.
Postel isn’t the only St. Joseph program head to have dramatically improved certain types of care at the hospital. Dr. Scott Sageman is an intensive care specialist who alternates weeks at St. Joseph and Mad River Community Hospital with pulmonary disease specialist Dr. Melvin Selinger. Sageman said interdisciplinary teams, standardized therapy and allowing nurses to “adjust on a minute-by-minute basis” now give St. Joseph the upper hand in the fight against ventilator-associated pneumonia in the intensive care unit.
Ventilator-associated pneumonia can develop up to 48 hours after intubation. Berwick’s Institute for Healthcare Improvement described VAP as the leading cause of death among hospital-acquired infections. Forty-six percent of intubated patients who come down with VAP die, compared with 32 percent of those who do not come down with it.
Each VAP adds days in intensive care and $40,000 to hospital costs, Berwick’s institute reports. St. Joseph reported 20 VAP cases 2006. Over the last 25 months: zero. Statistically, it’s a turnaround that saved 5.6 lives and $1.6 million.
“You need to ensure several interventions are done in every single patient, every single day, in exactly the same way,” Sageman said. “It is not as easy as it sounds.”
The interventions can be as easy as elevating the bed between 30 and 45 degrees to putting at the bedside endotracheal tubes to continuously suck out the bacterial build-up in the mouth and back of the throat.
“They’ve been able to instill processes in infection control that led to dramatic quality improvement,” Glaseroff said. “The challenge is to extend that across all the processes.”
Ploss said the teamwork and processes are a “work in progress” elsewhere in the hospital. “That system, the framework to deal with those kinds of processes — the checklist, the forms to make sure it’s known exactly what medications a patient received, that the discharge instructions were given and understood — that’s what still needs to be refined,” Ploss said about heart failure care. “We only started to do this in the last year.”
Starr said the effort to standardize processes and decrease potential errors by using tools such as order sets began 18 months ago.
Also part of the process: Applying the Lean Toyota Production System for automobiles to the hospital. Toyota is a worldwide leader in reducing waste of all sorts to its production plants to decrease costs, improve quality and increase satisfaction among customers and its workforce.
The Lean Toyota model has gained traction among health economists and hospital leaders. Both groups see hospitals in need of major transformation. One school of thought — discussed at the inaugural meeting of the Society of Health Economists in 2006 — is that since much of hospital care is an assembly line process, utilizing the Toyota model could bring “new eyes” to an industry that is far from lean.
One example of becoming leaner at the hospital: bundling equipment to save time, which is one of the areas of waste identified by Toyota.
“The hospital is redesigning how central line kits are put together, bundling the materials with gowns, masks and gloves so they can be more quickly accessed in an emergency,” Dr. Ellen Mahoney said last week. She is the hospital’s director of perioperative services. “When things have to be done quickly, you don’t want to put the patient in danger by making the physician hunt for gloves or go without.”
Central line infections come from the IV catheters placed close to a patient’s heart. Starr and Sageman report no central line infections in the last 13 months after having five in 2006.
“The goal is zero complications,” Sageman said.
“We strongly believe the[Medicare heart failure] data does not reflect the high quality care we provide for our heart failure patients,” stated Dennis Leonardi in an e-mail last week. As the chairman of the board of directors for St. Joseph and Redwood Memorial hospitals, he warned that statistics can be “misleading.”
Berwick isn’t buying it. He’s seen mortality vary by 400 percent between hospitals and says it’s “extremely hard” to be called an outlier.
“Everybody knows there are a lot of errors in hospital data,” he said. “When you measure mortality, some of that goes away. You can’t debate whether a patient is dead or alive.”
While Sageman called documentation for “external review” — aka, the public — a “steep learning curve” that may take awhile for some to learn, Shah had little patience for poor coding or chart documentation.
“That falls into the category of not doing a good job,” he said.
Berwick said being signaled out as “worse” for being more than two standard deviations from the mean should be thought of as a surprise — “the study authors going ‘Whoa. We didn’t expect this. There’s something going on.’”
It’s not about feeling good or bad about the mortality estimate, he said. It’s about learning. “If St. Joseph or any hospital says let’s work on X, they will make progress,” he said.
St. Joseph leadership expects more than progress. Leonardi and Starr both spoke of the hospital’s goal of “perfect care.” Depending on who is asked, the goal is worthy, unattainable, arrogant or evidence of a leadership that focuses more on platitudes and public assurances of quality than reality and measurable outcomes based on solid data.
“I know the hospital is trying, but I’m not sure it is accomplishing what it wants with patient care,” said long-time nurse Lavon Divine-Leal.
Nurse Kathryn Donahue shares those concerns. She points to staffing shortages and excessive paperwork as barriers to good care.
“The hospital will say it staffs to acuity, but I have a pile of complaints from nurses that it staffs to number, not acuity,” she said last week. “Every nurse on med-surg has five patients, no matter how sick they are. Staffing almost always impacts patient safety.”
Berwick said patients will not be the force for change.
“The mechanism that works is the super egos of the hospitals and the leaders,” he said. “They want to do well, to be the best, not the worst.”
If that’s the case, then the hospital and its board should be motivated by more than the heart failure numbers. Late last month, the Office of Statewide Health Planning and Development identified St. Joseph’s risk-adjusted inpatient mortality estimate for acute stroke treatment as worse than average for 2007.
OSHPD reported 16.9 percent of St. Joseph’s 96 acute stroke patients died, compared to 10.4 percent across California. That makes the hospital even more of an outlier in stroke care than in heart failure.
“We haven’t yet had a lot of time to analyze and interpret the stroke data,” Starr said. “Right now we’re reviewing each individual case with medical staff leadership to see what kind of improvements we can make.”
Ultimately, the medical leadership reports to the community folk who are board members today, but could be patients in the hospital tomorrow.
“I think you are seeing boards trying to wake up to their duty,” Berwick said. “It’s about conscience, purpose, commitment and leadership, not just money.”