The true horror of modern hospital medicine is starting to be revealed.<\/p>\n
440,000 deaths per year (up from 96,000 based on 1984 data) – one sixth of all deaths nationally, making preventable hospital error the third leading cause of death in the United States.<\/p>\n
http:\/\/opinionator.blogs.nytimes.com\/2013\/12\/04\/to-make-hospitals-less-deadly-a-dose-of-data\/?hp&rref=opinion&_r=1<\/a><\/span><\/p>\n Going to the hospital is supposed to be good for you. But in an alarming number of cases, it isn\u2019t. And often it\u2019s fatal. In fact it is the most dangerous thing most people will do.<\/p>\n Until very recently, health care experts believed that preventable hospital error caused some 98,000 deaths a year in the United States \u2014 a figure based on 1984 data. But a new\u00a0report<\/a>\u00a0from the Journal of Patient Safety using updated data holds such error responsible for many more deaths \u2014 probably around some 440,000 per year. That\u2019s one-sixth of all deaths nationally, making preventable hospital error the third leading cause of death in the United States. And 10 to 20 times that many people suffer nonlethal but serious harm as a result of hospital mistakes.<\/p>\n Most of us decide which hospital to go to (that is, when we get to decide) with zero data about hospital safety. Information, however, is gradually reaching the public, and it can do more than just help us choose wisely. When patients can judge hospitals on their safety records, hospitals will become safer. Just as publishing health care prices will drive them down, publishing safety information will drive hospital safety up.<\/p>\n In theory, finding this information shouldn\u2019t be a problem. Hospitals began to track errors seriously around 2000. The federal government\u2019s Centers for Medicare and Medicaid Services began collecting information on hospital quality in 2003, and since 2005 has been posting information on the website\u00a0Hospital Compare<\/a>. Many states have their own websites.<\/p>\n Other organizations compile this information as well, such as Consumers Union\u2019s\u00a0Consumer Reports (subscription required)<\/a>, which scores hospitals on their safety and the quality of care.\u00a0The Leapfrog Group<\/a>, which represents employer purchasers of health care, scores hospitals on\u00a0safety measures<\/a>. (The hospital ranking site probably most familiar to readers,\u00a0U.S. News\u2019 Best Hospitals rankings<\/a>, describes its mission as a very different one \u2014 to help patients with very difficult problems choose hospitals.)<\/p>\n All of these groups\u00a0measure different things<\/a>, which is why a hospital can rank near the top on one list and near the bottom on another. Most groups make money by charging hospitals to use their logo and ratings in their publicity. Consumer Reports is an exception \u2014 it doesn\u2019t allow hospitals to advertise its rankings.<\/p>\n \u201cThere is no longer a question of whether or not people have a right to information about quality, and that hospitals should be transparent and accountable,\u201d said Debra L. Ness, the president of the National Partnership for Women and Families. Ness is on the board of the\u00a0National Quality Forum<\/a>, the organization that sets standards for evaluating health care safety and quality. \u201cIt\u2019s not so much any longer a debate about whether \u2014 it\u2019s more about how.\u201d<\/p>\n But so far, the answer to the question of how is \u201cslowly.\u201d There is a big advance coming \u2014 Hospital Compare plans to begin reporting on rates of MRSA (or methicillin-resistant Staphylococcus aureus, a drug-resistant bacteria) and C-diff (Clostridium difficile) infections this month. These are dangerous, high-prevalence infections \u2014 crucial safety issues to track. But they are an exception on Hospital Compare. Much of what the public wants to know isn\u2019t there \u2014 and a lot of what\u2019s there isn\u2019t meaningful.<\/p>\n What\u2019s your hospital\u2019s rate of surgical site infection? You can find out if you live in\u00a0California<\/a>\u00a0or\u00a0Pennsylvania<\/a>\u00a0\u2014 states that collect exhaustive information on hospital infections and post it. The rest of us are out of luck. Hospital Compare will tell you only about colon surgery or abdominal hysterectomy \u2014 no knee replacement, heart bypass or any other surgery. How often does your local hospital leave a foreign object (like a surgical sponge) inside a patient? Or administer the wrong type of blood? Or allow a patient to develop a serious bed sore or a blood clot? Hospital Compare is now listing only old data for these errors, and has stopped updating those measures on the site.<\/p>\n What about the hospital\u2019s record at preventing re-admission in the 30 days after discharge? We can find that out for Medicare patients (the data comes from Medicare claims), but not for the rest of us. \u201cHospital Compare has a lot of bells and whistles but underneath it is nothing,\u201d said Leah Binder, the chief executive of The Leapfrog Group. \u201cMost hospitals are rated as average on every measure, and most measures are not things of great interest. We\u2019re further along, but we\u2019re really in the dark ages on reporting information in a way the public can use.\u201d<\/p>\n Measuring hospital safety is hard. Comparison, of course, requires everyone to be using the same measures \u2014 so how to reconcile the many variations hospitals use? And how do we know a measurement actually tells us what we think it does?<\/p>\n It\u2019s easiest to measure how often hospitals carry out processes that are recognized to be best practice, such as whether the patient got treatment to prevent blood clots after certain types of surgery, or whether the patient\u2019s temperature was kept steady in the operating room. Hospitals track such processes for their own internal quality controls.<\/p>\n This kind of process information dominates Hospital Compare and some of the independent rating organizations. (U.S. News\u2019 rankings lean heavily on a hospital\u2019s reputation, which earns it\u00a0heavy criticism<\/a>.)<\/p>\n But tracking processes doesn\u2019t produce the kind of information patients need. Hospitals are doing so well on these measures they are topping out, offering no way to compare them. Some of the measures are only loosely related to patient outcomes. For example, Hospital Compare shows that the national average for the practice of discontinuing prophylactic antibiotics within 24 hours after surgery is 97 percent. Top marks \u2014 but there is little evidence showing that this practice is linked to fewer surgical site infections. And it\u2019s outcomes that count.<\/p>\n Why doesn\u2019t Hospital Compare list more outcomes? Hospitals argue \u2014 and they are right \u2014 that it is much more expensive and technically difficult to develop outcome measures than process measures. \u201cWe need measures that have scientific reliability and validity,\u201d said Nancy Foster, the American Hospital Association\u2019s vice president of quality and patient safety policy. \u201cHospitals need the engagement of medical staff. If medical staff doesn\u2019t find the data credible then you lose them \u2014 they won\u2019t be there in the quality improvement. \u201c<\/p>\n But at times it seems as if hospitals aren\u2019t trying very hard. They like to report process measures on which they score well. But with 440,000 deaths from hospital error per year, their record is poor on key safety outcomes. This somewhat dampens their enthusiasm for public reporting. And what hospitals want matters a lot. \u201cAt the end of the day, the providers have to implement this,\u201d said Ness. \u201cThere has to be a reasonable amount of buy-in for it to work well.\u201d<\/p>\n \u201cIf you just looked at Hospital Compare\u2019s process measures, you\u2019d assume that all hospitals in this country are doing extremely well,\u201d said Binder. \u201cThis is misleading to the public because of the politics behind the scene of the website. Lobbyists for providers have been very effective at making sure what gets reported doesn\u2019t have much teeth.\u201d<\/p>\n Hospital Compare chooses what to display mainly using guidelines set by the National Quality Forum, which was established in 1999 in response to a government commission on consumer protection in health care. At the Quality Forum, groups representing health care consumers \u2014 patients and the corporations who pay for health care \u2014 are represented on all committees, and they hold a guaranteed majority on the most important committee. But patients can\u2019t match the clout of the providers. \u201cHospitals are ever-present in this work,\u201d said Lisa McGiffert, who is director of the Safe Patient Project at the Consumers Union and has been a consumer representative on several Quality Forum committees. \u201cThey have lobbyists all over Congress and administration folks. They outnumbered us on the committees that I have been on at N.Q.F. When I was on the infections committee I was rolled over constantly.\u201d<\/p>\n In a December 2011 meeting, the Measurement Application Partnership, a committee run by the Quality Forum, voted \u2014 over the objection of consumer and purchaser representatives \u2014 not to endorse reporting on several different serious hospital errors that were already on Hospital Compare. Hospital Compare then stopped updating data on air embolism, sponges or instruments left in a patient, serious bed sores and blood clots, among other events.<\/p>\n No one thought the raw data was unfair to hospitals \u2014 the data probably undercounted the number of hospital errors, said Foster. But hospitals argued that in some cases, the per-hospital numbers were so small the differences between hospitals might have been random, a conclusion supported by an independent review. (Hospitals have fought changes that would make reporting more complete \u2014 so it takes chutzpah to argue that the numbers are too small to publish.) \u201cWe agree with the concept,\u201d Foster said. \u201cBut the way the measures are executed makes them very unreliable and, we believe, invalid. You don\u2019t know that what you are looking at is an accurate representation of a hospital\u2019s performance.\u201d<\/p>\n Advocates for health care consumers argued that it didn\u2019t matter \u2014 just knowing the number of errors was important. \u201cDo you as an American have the right to know if the hospital down the street left an object in a patient?\u201d said Binder. \u201cThat information has now been taken out of the hands of the consumer by lobbyists. We should always tilt towards transparency.\u201d<\/p>\n Poor or irrelevant data keeps patients from finding the information they need. Another problem is that the data that\u2019s there isn\u2019t presented in a way people can easily use.<\/p>\n Hospital Compare cuts very thick slices. There\u2019s below average, above average and average, which is the score of the vast majority of hospitals. And most patients simply don\u2019t know about Hospital Compare. That\u2019s not the government\u2019s fault, but it does illustrate the need for translator organizations such as Consumer Reports \u2014 which has five categories, not three \u2014 and Leapfrog, which issues letter grades, with more detail available for those who want it.<\/p>\n Leapfrog\u2019s twice-yearly data release gets a lot of coverage. McGiffert said that when Consumer Reports first came out with ratings for central-line and surgical site infections, some hospitals protested that the data was wrong. But it was the same data hospitals had submitted for state reports. \u201cWe were using data that had already been on state websites, but nobody had paid attention to it,\u201d she said. \u201cAgencies are never going to do a media push when they publish these.\u201d<\/p>\n That media push reaches more patients, and it forces hospitals to focus on safety. \u201cThese are a major factor in getting hospitals\u2019 attention,\u201d said McGiffert. She said that hospitals in states that required public reporting were far more likely to adopt quality-improvement practices.<\/p>\n Binder said that except for advances in doctors\u2019 using computers to enter treatment orders, hospital safety records, as a group, are not improving. This is hardly surprising. What gets measured gets done, and many aspects of safety are still not even measured. The Journal of Patient Safety study found 210,000 \u201cdetectable\u201d deaths per year \u2014 the number they eventually fixed on of 440,000 reflected the estimate that half or two-thirds of all such deaths are never counted. \u201cThat\u2019s a big range,\u201d said Binder. \u201cIt sounds so high, but what more frightening is that we still don\u2019t know. Nobody\u2019s counting the bodies.\u201d<\/p>\n Join\u00a0Fixes on Facebook<\/a>\u00a0and follow updates on\u00a0twitter.com\/nytimesfixes<\/a>. To receive e-mail alerts for Fixes columns, sign up\u00a0here.<\/a><\/em><\/p>\n Tina Rosenberg won a Pulitzer Prize for her book \u201cThe Haunted Land: Facing Europe\u2019s Ghosts After Communism<\/a>.\u201d She is a former editorial writer for The Times and the author, most recently, of \u201cJoin the Club: How Peer Pressure Can Transform the World<\/a>\u201d and the World War II spy story e-book \u201cD for Deception.\u201d<\/em><\/p>\n<\/div>\nBy\u00a0TINA ROSENBERG<\/a><\/address>\n Going to the hospital is supposed to be good for you. But in an alarming number of cases, it isn\u2019t. And often it\u2019s fatal. In fact it is the most dangerous thing most people will do.<\/p>\n Until very recently, health care experts believed that preventable hospital error caused some 98,000 deaths a year in the United States \u2014 a figure based on 1984 data. But a new\u00a0report<\/a>\u00a0from the Journal of Patient Safety using updated data holds such error responsible for many more deaths \u2014 probably around some 440,000 per year. That\u2019s one-sixth of all deaths nationally, making preventable hospital error the third leading cause of death in the United States. And 10 to 20 times that many people suffer nonlethal but serious harm as a result of hospital mistakes.<\/p>\n Most of us decide which hospital to go to (that is, when we get to decide) with zero data about hospital safety. Information, however, is gradually reaching the public, and it can do more than just help us choose wisely. When patients can judge hospitals on their safety records, hospitals will become safer. Just as publishing health care prices will drive them down, publishing safety information will drive hospital safety up.<\/p>\n In theory, finding this information shouldn\u2019t be a problem. Hospitals began to track errors seriously around 2000. The federal government\u2019s Centers for Medicare and Medicaid Services began collecting information on hospital quality in 2003, and since 2005 has been posting information on the website\u00a0Hospital Compare<\/a>. Many states have their own websites.<\/p>\n Other organizations compile this information as well, such as Consumers Union\u2019s\u00a0Consumer Reports (subscription required)<\/a>, which scores hospitals on their safety and the quality of care.\u00a0The Leapfrog Group<\/a>, which represents employer purchasers of health care, scores hospitals on\u00a0safety measures<\/a>. (The hospital ranking site probably most familiar to readers,\u00a0U.S. News\u2019 Best Hospitals rankings<\/a>, describes its mission as a very different one \u2014 to help patients with very difficult problems choose hospitals.)<\/p>\n All of these groups\u00a0measure different things<\/a>, which is why a hospital can rank near the top on one list and near the bottom on another. Most groups make money by charging hospitals to use their logo and ratings in their publicity. Consumer Reports is an exception \u2014 it doesn\u2019t allow hospitals to advertise its rankings.<\/p>\n \u201cThere is no longer a question of whether or not people have a right to information about quality, and that hospitals should be transparent and accountable,\u201d said Debra L. Ness, the president of the National Partnership for Women and Families. Ness is on the board of the\u00a0National Quality Forum<\/a>, the organization that sets standards for evaluating health care safety and quality. \u201cIt\u2019s not so much any longer a debate about whether \u2014 it\u2019s more about how.\u201d<\/p>\n But so far, the answer to the question of how is \u201cslowly.\u201d There is a big advance coming \u2014 Hospital Compare plans to begin reporting on rates of MRSA (or methicillin-resistant Staphylococcus aureus, a drug-resistant bacteria) and C-diff (Clostridium difficile) infections this month. These are dangerous, high-prevalence infections \u2014 crucial safety issues to track. But they are an exception on Hospital Compare. Much of what the public wants to know isn\u2019t there \u2014 and a lot of what\u2019s there isn\u2019t meaningful.<\/p>\n What\u2019s your hospital\u2019s rate of surgical site infection? You can find out if you live in\u00a0California<\/a>\u00a0or\u00a0Pennsylvania<\/a>\u00a0\u2014 states that collect exhaustive information on hospital infections and post it. The rest of us are out of luck. Hospital Compare will tell you only about colon surgery or abdominal hysterectomy \u2014 no knee replacement, heart bypass or any other surgery. How often does your local hospital leave a foreign object (like a surgical sponge) inside a patient? Or administer the wrong type of blood? Or allow a patient to develop a serious bed sore or a blood clot? Hospital Compare is now listing only old data for these errors, and has stopped updating those measures on the site.<\/p>\n What about the hospital\u2019s record at preventing re-admission in the 30 days after discharge? We can find that out for Medicare patients (the data comes from Medicare claims), but not for the rest of us. \u201cHospital Compare has a lot of bells and whistles but underneath it is nothing,\u201d said Leah Binder, the chief executive of The Leapfrog Group. \u201cMost hospitals are rated as average on every measure, and most measures are not things of great interest. We\u2019re further along, but we\u2019re really in the dark ages on reporting information in a way the public can use.\u201d<\/p>\n Measuring hospital safety is hard. Comparison, of course, requires everyone to be using the same measures \u2014 so how to reconcile the many variations hospitals use? And how do we know a measurement actually tells us what we think it does?<\/p>\n It\u2019s easiest to measure how often hospitals carry out processes that are recognized to be best practice, such as whether the patient got treatment to prevent blood clots after certain types of surgery, or whether the patient\u2019s temperature was kept steady in the operating room. Hospitals track such processes for their own internal quality controls.<\/p>\n This kind of process information dominates Hospital Compare and some of the independent rating organizations. (U.S. News\u2019 rankings lean heavily on a hospital\u2019s reputation, which earns it\u00a0heavy criticism<\/a>.)<\/p>\n But tracking processes doesn\u2019t produce the kind of information patients need. Hospitals are doing so well on these measures they are topping out, offering no way to compare them. Some of the measures are only loosely related to patient outcomes. For example, Hospital Compare shows that the national average for the practice of discontinuing prophylactic antibiotics within 24 hours after surgery is 97 percent. Top marks \u2014 but there is little evidence showing that this practice is linked to fewer surgical site infections. And it\u2019s outcomes that count.<\/p>\n Why doesn\u2019t Hospital Compare list more outcomes? Hospitals argue \u2014 and they are right \u2014 that it is much more expensive and technically difficult to develop outcome measures than process measures. \u201cWe need measures that have scientific reliability and validity,\u201d said Nancy Foster, the American Hospital Association\u2019s vice president of quality and patient safety policy. \u201cHospitals need the engagement of medical staff. If medical staff doesn\u2019t find the data credible then you lose them \u2014 they won\u2019t be there in the quality improvement. \u201c<\/p>\n But at times it seems as if hospitals aren\u2019t trying very hard. They like to report process measures on which they score well. But with 440,000 deaths from hospital error per year, their record is poor on key safety outcomes. This somewhat dampens their enthusiasm for public reporting. And what hospitals want matters a lot. \u201cAt the end of the day, the providers have to implement this,\u201d said Ness. \u201cThere has to be a reasonable amount of buy-in for it to work well.\u201d<\/p>\n \u201cIf you just looked at Hospital Compare\u2019s process measures, you\u2019d assume that all hospitals in this country are doing extremely well,\u201d said Binder. \u201cThis is misleading to the public because of the politics behind the scene of the website. Lobbyists for providers have been very effective at making sure what gets reported doesn\u2019t have much teeth.\u201d<\/p>\n Hospital Compare chooses what to display mainly using guidelines set by the National Quality Forum, which was established in 1999 in response to a government commission on consumer protection in health care. At the Quality Forum, groups representing health care consumers \u2014 patients and the corporations who pay for health care \u2014 are represented on all committees, and they hold a guaranteed majority on the most important committee. But patients can\u2019t match the clout of the providers. \u201cHospitals are ever-present in this work,\u201d said Lisa McGiffert, who is director of the Safe Patient Project at the Consumers Union and has been a consumer representative on several Quality Forum committees. \u201cThey have lobbyists all over Congress and administration folks. They outnumbered us on the committees that I have been on at N.Q.F. When I was on the infections committee I was rolled over constantly.\u201d<\/p>\n In a December 2011 meeting, the Measurement Application Partnership, a committee run by the Quality Forum, voted \u2014 over the objection of consumer and purchaser representatives \u2014 not to endorse reporting on several different serious hospital errors that were already on Hospital Compare. Hospital Compare then stopped updating data on air embolism, sponges or instruments left in a patient, serious bed sores and blood clots, among other events.<\/p>\n No one thought the raw data was unfair to hospitals \u2014 the data probably undercounted the number of hospital errors, said Foster. But hospitals argued that in some cases, the per-hospital numbers were so small the differences between hospitals might have been random, a conclusion supported by an independent review. (Hospitals have fought changes that would make reporting more complete \u2014 so it takes chutzpah to argue that the numbers are too small to publish.) \u201cWe agree with the concept,\u201d Foster said. \u201cBut the way the measures are executed makes them very unreliable and, we believe, invalid. You don\u2019t know that what you are looking at is an accurate representation of a hospital\u2019s performance.\u201d<\/p>\n Advocates for health care consumers argued that it didn\u2019t matter \u2014 just knowing the number of errors was important. \u201cDo you as an American have the right to know if the hospital down the street left an object in a patient?\u201d said Binder. \u201cThat information has now been taken out of the hands of the consumer by lobbyists. We should always tilt towards transparency.\u201d<\/p>\n Poor or irrelevant data keeps patients from finding the information they need. Another problem is that the data that\u2019s there isn\u2019t presented in a way people can easily use.<\/p>\n Hospital Compare cuts very thick slices. There\u2019s below average, above average and average, which is the score of the vast majority of hospitals. And most patients simply don\u2019t know about Hospital Compare. That\u2019s not the government\u2019s fault, but it does illustrate the need for translator organizations such as Consumer Reports \u2014 which has five categories, not three \u2014 and Leapfrog, which issues letter grades, with more detail available for those who want it.<\/p>\n Leapfrog\u2019s twice-yearly data release gets a lot of coverage. McGiffert said that when Consumer Reports first came out with ratings for central-line and surgical site infections, some hospitals protested that the data was wrong. But it was the same data hospitals had submitted for state reports. \u201cWe were using data that had already been on state websites, but nobody had paid attention to it,\u201d she said. \u201cAgencies are never going to do a media push when they publish these.\u201d<\/p>\n That media push reaches more patients, and it forces hospitals to focus on safety. \u201cThese are a major factor in getting hospitals\u2019 attention,\u201d said McGiffert. She said that hospitals in states that required public reporting were far more likely to adopt quality-improvement practices.<\/p>\n Binder said that except for advances in doctors\u2019 using computers to enter treatment orders, hospital safety records, as a group, are not improving. This is hardly surprising. What gets measured gets done, and many aspects of safety are still not even measured. The Journal of Patient Safety study found 210,000 \u201cdetectable\u201d deaths per year \u2014 the number they eventually fixed on of 440,000 reflected the estimate that half or two-thirds of all such deaths are never counted. \u201cThat\u2019s a big range,\u201d said Binder. \u201cIt sounds so high, but what more frightening is that we still don\u2019t know. Nobody\u2019s counting the bodies.\u201d<\/p>\n Join\u00a0Fixes on Facebook<\/a>\u00a0and follow updates on\u00a0twitter.com\/nytimesfixes<\/a>. To receive e-mail alerts for Fixes columns, sign up\u00a0here.<\/a><\/em><\/p>\n Tina Rosenberg won a Pulitzer Prize for her book \u201cThe Haunted Land: Facing Europe\u2019s Ghosts After Communism<\/a>.\u201d She is a former editorial writer for The Times and the author, most recently, of \u201cJoin the Club: How Peer Pressure Can Transform the World<\/a>\u201d and the World War II spy story e-book \u201cD for Deception.\u201d<\/em><\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":" The true horror of modern hospital medicine is starting to be revealed. 440,000 deaths per year (up from 96,000 based on 1984 data) – one sixth of all deaths nationally, making preventable hospital error the third leading cause of death in the United States. http:\/\/opinionator.blogs.nytimes.com\/2013\/12\/04\/to-make-hospitals-less-deadly-a-dose-of-data\/?hp&rref=opinion&_r=1 To Make Hospitals Less Deadly, a Dose of Data DECEMBER … Continue reading To Make Hospitals Less Deadly, a Dose of Data<\/span> To Make Hospitals Less Deadly, a Dose of Data<\/span><\/h3>\nDECEMBER 4, 2013,\u00a011:00 AM
\nBy\u00a0TINA ROSENBERG<\/a><\/address>\n