Experts estimate that hundreds of thousands of patients die each year from an error, injury or infection sustained in a hospital. Many more are harmed or receive unnecessary treatment. We’re curious:

A) Do you really believe that this many patients suffer some type of harm? Why or why not?

B) Who or what do you think is responsible for the ongoing patient harm?

We've invited several health care providers to answer.
Our reporters Marshall Allen and Olga Pierce, and Senior Editor Tom Detzel are also here.

Hey everyone, thanks for participating in our chat. We're excited to see how this conversation proceeds.

a) As an anesthesiologist, yes, I do believe those numbers are roughly correct. Any time a patient enters the hospital they are assuming risk. Hospitals are not 'miracle centers'. They are complex systems in which patients have almost no defenses with which to protect themselves against an unfamiliar environment and multiple dangers (reduced reserve, drugs, germs, procedures, cost-saving efforts, production pressure, for example).

b) Much tougher, of course. Part of the answer is patients themselves. Much higher obesity rates, continued high smoking rates (in my part of the country anyway). Those two items themselves increase wound infection rates, and make it more difficult to do almost anything you can think of in the hospital.

I don't challenge the numbers, but I do feel like they might be invoked sensationally at times. I have never felt good about the frequent use of the 747 image. Scaring patients into thinking their clinicians might be incompetent can have its own adverse effects. That said, there is plenty of room to improve, and much of it is OUTSIDE the hospital.

Who to blame? I would say payment system that does not incentivize teamwork and value (outcomes/cost).

It's "only" a 727, Tom. But the image is powerful because if that level of harm were to occur, we'd close down the airports after about 3 days to fix it. The dispersed nature of harm in hospitals is different, notably in the lack of concerted effort on the part of the profession to solve it.

Incentives have little to do with this. It is fundamentally a flaw in how MDs are trained, with a lack of attention to how work is organized, systemic problems, and applying the scientific method to clinical process improvement. The Lucien Leape Institute noted in 2010:

“Medical schools are not doing an adequate job of facilitating student understanding of basic knowledge and the development of skills required for the provision of safe patient care.”

A. would be easier to address if the number of adverse outcomes were compared to the number of encounters i.e. out of every 100 encounters there are 99 or .9% negative outcomes. A higher acuity person will be more vulnerable to infection, I guess A needs more explanation. I do believe that there are many people exposed to negative outcomes, errors, injuries and infection.
B. To general. Errors:education, injuries:investigation and the cause addressed, It is almost impossible to control exposure to infectious organism in a hospital esp. with at risk patients, a good place to start is greater emphasis on hand washing and clean environment by all staff and visitors. Organisms are passed by everyone not just that other person.

Hi CLKeyes, a review of medical records by physicians, on behalf of the Office of Inspector General, found that one in seven Medicare patients suffered harm: propublica.org

The doctors who did the review estimated that about half of the harmful events were preventable: propublica.org

An estimated 1.5 percent of Medicare patients suffered a harmful event that contributed to their death, which projects to about 15,000 patients in a single month: propublica.org

I'm curious about the training issue: is patient safety training widespread? What are the skills that are/should be taught? Are administrators and/or providers receptive to such training?

a. There is no question that patients are being harmed due to safety problems in health care delivery. I and others have examined failure to rescue - death after a postoperative complication. Not only are the rates unacceptably high, some hospitals are much better are rescue than others. This is only the tip of the iceberg. Many patients are discharged from the hospital only to acquire a problem after their stay. It is challenging with our current data systems to examine whether this is attributable to their care or not.

b. It's multi-factorial. Providers fail to report "near-misses" that give us best chance to learn and correct problems. Patients and families may assume complications are expected and nothing to worry about.

Those numbers may be an underestimation. Some level of harm is unavoidable given the complexity of health care systems and the conditions being treated. Although any harm is bad, perhaps the "worst" form is harm related to unnecessary care, usually in the form of procedures performed for profit rather than to improve quantity or quality of life.

M. Allen, thank you. Nursing training re: safety. In school we learn physiological process, treatments, outcomes and care, within ea. category safety would be addressed. In hosp. experience included, annual safety training re: equip., infection control, chem., fire safety and CPR. Accreditation of hosp. q3 years: review and assessment of incidences, standards of care, policy updating. These areas were across the board for all levels of bedside care providers. In addition, each event would require documentation, reporting and follow-up training re: event. Variables effecting outcomes; lower staffing ratios, higher usage of CNA, LPN, tech staff vs. RN, shorter hosp. stays resulting higher acuity at discharge/ higher acuity on admit.

Administrators are amenable to training if it will make the outcomes look better and certainly in my experience there were ongoing training sessions provided. Care provided by nursing has detailed written procedures i.e. IV, catheters, medication, equipment use, dressing changes, monitoring parameters. I would have to say there isn't anything that nursing does that doesn't have a procedure. We are trained and tested, we have check lists and algorithms. In my experience errors occurred more frequently when there is not good communication between MD/RN staff. Nursing education at times focuses too much on theories and passing the St.Boards and not on the basic science needed for the critical thinking needed at the bedside.

The article Marshall Allen cites (propublica.org) is based on a review of 780 charts and appears to me to be an attempt to validate a method for determining adverse events based on retrospective chart reviews, then extrapolate. I applaud the effort, and knowing how poor charting is at our hospital, feel this must be an UNDER estimate. For example, just today an IV infiltrated on induction of anesthesia in an 80 year old woman with COPD and very poor veins. I noted it in the anesthetic record, but would they have looked there? My point is that the OIG report, rather than authoritative, represents one attempt to quantify using what is to them a new methodology. It would be interesting to do a 'synchronous' adverse event analysis.

Olga, Training in quality and safety improvement, and related training in disclosure and apology, is notably missing from medical schools and even residency programs. The only longitudinal patient Q&S program of which I am aware was developed by David Mayer and Tim McDonald at Univ. of Illinois Chicago Medical School.
The ACGME now requires training in these areas as part of residencies, but this is observed in the breach more than in rigorous and pedagogically appropriate programs.
This blog by Mayer and friends covers these topics in some depth and is worth reading: educatetheyoung.wordpress.com

My question after reading the report cited was, is 780 a large enough sample to extrapolate and come up with "hundreds of thousands of people dying from errors". I have done retrospective chart reviews and they are tremendously hard to pick out the little details needed to complete a survey. I am not in any way suggesting that these errors don't occur but I think the numbers seem high in light of my experience. I would have to say that Dr. Venable's infiltrated IV would be, in my experience much more common than a wrong pt/site/limb surgery. My background includes Surg. ICU, Recovery Room, and Day surgery and know the detailed process a pre-operative patient went through where I worked. However, the major question is what can be done?

In many hospitals quality and safety improvement have become stand alone departments and are specialties in themselves. However most are manned by nursing and administrative staffs, occ. a physician is part of the teams. I think that the lack of formal education before starting in the clinical setting in these areas puts doctors and nurses at odds when in high pressure situations and written procedures appear to hinder the productivity. When you are very busy the crossing of T's and dotting of i's feel like burdens if they are not ingrained from the beginning.

Olga: Nursing has responded to the call to develop Quality and Safety Education for Nurses (QSEN.org). With funding from the Robert Wood Johnson Foundation, we have developed core competencies for nurses in training, and also continuing education. Schools of Nursing and hospitals are now trying to integrate this content into an already-stacked curriculum.

Medical errors are understood relatively well, but we approach them like the rest of the health care system: piecemeal, scattered. The Keystone program has showed us how ridiculously simple improving patient safety can be.

See whitecoatunderground.com

We need to understand that medicine is not the same as usual market-driven industries: it requires an amount of top-down control, somewhat like commercial airlines, where unifying standards of safety create systems where errors are less likely (always inevitable, but less likely).

Errors are understood and as the OIG report shows many preventable. The solutions have been two fold, increased documentation and increased education at the source. Where I see the problems is the the top-down control. Give an the admin. a list of errors and they will create a committee, 3 binders of policies and mandatory education and give themselves a raise. Nursing is hands on, intensive patient contact, and has developed into documentation process. As costs increase "boots on the ground" staff has decreased, but the ways to document productivity has not. Computerization at times has increased work not decreased. Closer observation at the bedside by those who understand what they are seeing would decrease considerably the errors.

I have had the opportunity to be involved in patient safety issues at my institution. I believe systems approaches are valid; however, there are no consequences for providers' providing unsafe care. As long as patient harm does not translate into direct, measurable impact for the institution or the provider in terms of lost revenue or volumes, we may be fighting a losing battle. Transparency and clear reporting of outcomes are essential. Even Medicare with it's new Value Based Purchasing initiative only withholds 2% in 2017 We fire people when there are obvious systems failures (e.g. the transplant folks in Pittsburgh) but there are no sequelae for unsafe care. How can we systematize, standardize and mandate this type of reporting?

whitecoatunderground.com discusses check lists re: IV insertion, 2009. the precautions listed were not new in 2009 and if we are still discussing these issues now, more system studies, reporting and measuring is not going to decrease the errors. I helped develop check-lists, pre-admit, anesthesia, discharge instructions, policies as well as working the units. We have had "incident reports" since stone tablets, however with every documentation, "talking to", updating the procedures, training staff, and new mandated report time is taken away from the nurse/staff who would check that IV again, double check the vs, is that ET tube secured, is this the right pt., right procedure, right medication, does the pt. understand exactly what is being done?

the success of keystone is the systemization of checklists, akin to airlines, and giving the nurses power to halt a non-emergent procedure.

It is and does work, not disagreeing at all. Have created them and used them. An important part of check lists is that everyone in the the chain has to agree to work with them. "I don't need no stinkin checklist" or we don't have time for checklist protocols are a sure to lead back to the errors or neg. outcomes. I do think that with increased physician involvement and "buy-in" the success rate has/will increase. There is nothing more harmful than the medical staff's disdain for procedures that they have judged "wasteful and necessary". The next step is to instill a solid sense of "take the time to do it right" in the staff and the courage to speak up when the checklist chain is broken.

Right, CL. As Captain Sullenberger notes, "A checklist alone is not sufficient. What makes it effective are the attitude, behavior and teamwork that go along with the use of it." A checklist applied to an organization that has not engaged in a cultural change that promotes respectful relationships among the clinical staff--described by Peter Pronovost as a comprehensive unit-based safety program--will fail. Peter’s contribution in his book Safe Patients, Smart Hospitals is explaining how to create such an environment and in telling the story of his successes and failures along the way. More here: runningahospital.blogspot.com

PL why are we still struggling with the barriers between outcomes and knowledge of causes. It seems to me that we keep talking about the reasons why and how, but that doesn't trickle past the words on the page. Is it flaws in the education of staff, the quantity of staff, or the commitment of individual staff to do the very best job they can? I am prejudiced in this question because I think that it is a top down problem and that we don't need more studies or "solutions" we need to put hands on the job. How many of the errors in the OIG would have been solved it the nursing staff were at the bedside and not at the desk documenting why they needed that nurse at the bedside? The is true across the continuum of care, not just acute care.

The staff people are fully committed, CL. Usually, the leadership does not know how to coach/teach/lead them in clinical improvement methods, redesigning work flows, front-line driven process improvement. Some places have done it. Among my favorites are Gundersen Lutheran in La Crosse, WI, where CEO Jeff Thompson, his board, and clinical leaders get it and do it. There are a few others, too. What makes them different? Intellectual modesty and a commitment to full-scale staff engagement, drawing on the value system that exists for every MD, RN, and others. It's not so hard if the leaders really want to make a difference. And, btw, it has NOTHING to do with financial incentives. That is a distraction, not motivational to the staff.

I agree financial incentives are a distraction for staff and definitely not motivational if you are part of the "front-line" staff. However I do believe that the higher up the administrative ladder you go that the financial aspects becomes a bigger part of the picture. The example of betadine vs chlorhexidine is excellent, but an element that was not mentioned is that many kits are pre-made and purchased from major suppliers, so the contents of the kits need to be adjusted and there is where the financial incentive steps in, is a betadine kit cheaper, is there a better contract with that supplier than the other? Does the purchasing dept. appreciate the need for change, does the manufacturer gain anything in the change?

Since we know what, how many and why, what is the next step? Are these errors regional? The sample is pretty small so that might be hard to determine, but I am sure there are reports out there somewhere with that answer. A mobile team of medical/nursing staff assigned to a facility with a certain percentage of errors could evaluate/teach and facilitate changes. Would those costs be greater than the resulting costs in longer stays, liability coverage/compensation, and continual losses RT to poor outcomes? There are a plethora of accreditation and quality products out there, but it seems that we would need more of a sniper approach as opposed to the shot gun approach, since gathering data doesn't equate to improvement.

I think a key challenge is "Quality Fatigue." Clinicians are deluged with run charts prepared by well-intentioned people. They don't know how to act on the information. It takes leadership to say, "we value your suggestions to improve," and "how can we as leaders streamline processes so you can focus on the right things?" I'm continually amazed at physicians having to handle clerical matters and nurses handle errant food trays. We need to let clinicians be clinicians, give them the tools and resources to make good decisions, and track how we do over time.

This discussion is largely around what I would call "surgical misadventure" which is something that can be usefully addressed with checklist, and has been in many hospital systems. But this is not the major cause of medical error...which is medication errors and adverse drug reactions. These things cause huge harm that is often recognized by patients, but rarely by providers and only on the very severe sense by healthacre systems..i.e. when they get sued. We need systems that educate provider much better about ADRs an how to recognize and report them and that train doctors specifically better in therapeutics

Hospital medication check lists on admission and discharge for drugs are minor administrative musts that have been in place for years but teh system of verbal prescribing for prescription medicines is a huge problem, when this could be digitized fairly easily and made to work on smart phones, which do not confuse "Lasix" with "Losec" or get doses wrong etc. Drug interactions are also a particularly bad source of egregious errors. Alert fatigue created by the standard pharmacy alerts systems and a gross lack of training in how to handle even bad interactions are big system problems.

I agree about buy-in. Docs at my hospital responded negatively to keystone because no one told them anything about it, especally those who would be exposed to it less often. It was a real failure of leadership.

Also, along with quality fatigue, the way medicare and jchao requirements are implemented is insane at some institutions. OUr discharge instructions are un-readable for patients, and contain useless information (if...then's, etc). I have yet to see pt at a post op visit who understood their d/c instructions.

Hospital and outpatient settings, and NH settings are very different as well, and require different standards. And while we're on it, blatant quackery has become a real distraction. Many of my patients take all kinds of barely-regulated useless potions making it harder to insure that the drugs I do give them are being used safely.

PalMD do you have links for your Keystone project I am not familiar with it. I have to agree about the jchao requirements. It has always been my contention that the changing everything around every 3 years was because it gave them jobs and they had a state in the paper industry. The staff time in cost and productivity has to be enormous and yet we still see the list of errors and complaints that haven't changed over the years. DAFlockhart are there satellite pharmacies and pharmacists at your facilities, is there evidence if that has decreased medication errors? I can see EMR use for inputting medication orders that would stop the order if the dose/med was incorrect and or there were interaction issues.

I think i linked the data in my blog post.

There are satellite pharmacies and lots of pharmacists around an academic medical center like this, and yes there are data indicating that they can reduce medication errors, so thats worth asking about if you're ever admitted.That said, the effect of pharmacist presence is real, but fairly small.Bad stuff still happens

EMR should be an incredibly valuble tool for reducing medication errors and deleterious drug interactions. It cannot help us anivipate ADRs, but it can help us manage them and mitigate their effects

A brief review of articles, many dated back in the 1990's.(dramatic headlines in CNN, criteria by the WHO, BMJ, Harvard study etc.) It seems that the same general complaints continue today. Wrong patient, wrong medication, the nurse didn't this/that, too much pain med, not enough pain med.,too many to list here,these are issues that we have been studying since I started in nursing in the 1970's. Have we really made so little progress in these areas? It is bewildering that we are still doing the same things wrong. Feeding tubes and IV lines have never looked alike or are seldom in the same part of the body, nurses/MD in surg. can read and count, electronic devices abound. Do we need to stop gathering data and start looking at solutions?

I'll speak to a particular patient-safety problem, with the hope that my experiences might point to broader lessons
for patient safety. I have been pursuing the issue for more than 25 years, from the perspective of a hospital nurse.

The problem I saw was pretty simple: defibrillation took too long in cases of in-hospital cardiac arrest. The equipment (defibrillators) was generally available, but trained users were not. The remedy: train staff likely to be near the scene of an arrest (i.e., nurses) to defibrillate.

The American Heart Association acknowledged the problem of delayed defibrillation, but pursued a course that was lucrative for defibrillator manufacturers but proved to be ineffective in improving survival.

More to come.

That's interesting; i'd imagine it varies from hospital to hospita. I always felt like our code team was timely, but i suppose slapping on an AED before anyone gets there isn't the worst idea. However:
most in-hospital arrests don't end well, and not because of response time. Patients who code in hospital tend not to be cases of sudden cardiac death but complications of medical or surgical conditions that are harder to recover from.

I did not really mean to get into talking about the problem itself--how widespread or serious it is, possible remedies, etc.--in this forum, which I think is intended to be a more general discussion. I thought the history of a patient-safety problem--one that has been recognized as serious by the AHA but has not been remedied after many years--might provide some general lessons about the change process in healthcare.

That having been said, I am more than willing (after all these years) to discuss at great length and in excruciating detail why this is a major (costing tens of thousands of lives in the US alone) and solvable patient-safety problem. Maybe that is appropriate for the Branch feature here, if I can figure out how to use it.

I dont think codes are a significant patient safety issue, compared to med errors, nosocomial infections, etc. Most of these are preventable with the use of information technology and proper systems of checklists, and by simple to enforce rules about, say, handwashing. It helps to empower non-physician staff to enforce this, but all this must be properly communicated to doctors, which is rare IME.

I'll move on from the question of the significance of delayed in-hospital defibrillation as a patient-safety issue, hoping to create a "branch" discussion about it soon. But rightly or wrongly, the tremendously influential American Heart Association, source of the widely (really almost universally) accepted Emergency Cardiac Care Guidelines, has recognized it as a significant problem. I want to cast some light on how the AHA came to that position, how they came up with a proposed remedy in-hospital use of AEDs), and why that remedy failed.

While in graduate school in the mid-80s ( instructional design; thesis project concerned training nurses to defibrillate), I brought the issue to the attention of several AHA officials.

More to come.

The officials I approached on the issue in grad school were generally supportive of my efforts. After grad school, I moved to Seattle hoping to form a collaboration with, or at least get some leads from, an emergency medicine MD who had been supportive. However, he gave me the brushoff.

A couple of years later, that MD and the other AHA people I had alerted to the problem began presenting and publishing about it, with no acknowledgment to me for raising the issue--and more importantly, fixed on a "solution" without public discussion. Their remedy was purchasing AEDs for use in hospitals. This was ill-advised for several reasons, but it was certainly lucrative for the defibrillator industry. I dissented, both privately and in print.