Thursday, February 12, 2009

Good and bad news about infection control

I have been writing for some time about our efforts to eliminate central line infections in our hospital, and we have been totally transparent about our progress in that regard. While I know you can always look these things up, I want to make it easier for you and give you some advance news -- especially in light of the most recent results.

During the first four months of this fiscal year, a period covering about 7000 patient days, we had only one CLI in our intensive care units. This represents a tremendous effort by dozens and dozens of staff people.

In early 2006, our hospital's rate of infections was about 2 per thousand ICU patient days. At that old rate, there would have been 14 infections during this same four-month period. Given a 12 to 25 percent mortality rate associated with such infections, 2 or 3 people would have died unnecessarily.

Do we need a better reason to engage in these programs?

Two years ago, I raised a question: "If I can post these rates for BIDMC, why can't people from other hospitals? ... I am seeking no competitive advantage here. This is an attempt to get past a culture of blame and litigation and persuade people that transparency works: Real-time public disclosure of key indicators like this ... can be mutually instructive and can help provide an incentive to all of us to do better."

Then, a short while ago, I asked the question in a more direct way, posing a challenge to all the Boston area hospitals to jointly engage in a program to eliminate these kinds of infections and share their progress with the public.

The response to my public and private entreaties in this realm has been silence -- from hospital professionals, from insurance executives who care about a transformation of this industry, and, indeed, from public advocacy groups who care about access to care and the quality of care delivered. Some observers attribute the medical profession's lack of engagement to an underlying fear of transparency. And yesterday, a world expert in this field, whose wisdom and advice I treasure, told me that he has come to accept gradual progress in quality and safety improvement, citing the kind of training doctors get, which does not emphasize these areas. That such a person has become content with gradual changes in the status quo is an indication of what it must be like to beat your head against this wall of recalcitrance for several decades.

My advantage, being without medical training and having had but a short tenure in this field, is that I retain a sense of outrage. Our collective failure to approach this problem using well established methods of process improvement -- including publication of current performance results -- represents a moral and ethical lapse by the clinical and administrative leadership of the medical establishment in this city. Why? Simply put, a profession that takes an oath to do no harm is, by inaction or incomplete action, doing harm. We are causing people to die who should not die. What would we call that if we saw it happening in other sectors of society?


AnnB said...

You think that kind of recalcitrance is bad? You should try working in Ireland, patients and doctors are issued with crash helmets when we enter the health system such is the brick wall mentality. I commend your approach and wish you luck, but,if I may offer advice, as a veteren of 5 years battling for my son, logic, ethics and common sense often have no place in large health care systems. Those of us who operate on that basis sometimes get felled in the field of combat.
Keep up the great work I wish you were running our hospital!

Anonymous said...

Very Good. Keep up with the good work. Quality Improvement in IPC requires a certain level of humility. To be able to say this is what we are doing wrong and this is how bad the situation is, and committing to improve.

It always helps when the facility manager has your kind of passion.

Yes, HCW's do have fears. Until such time that we can change legislation and regulations and align them with the just culture, fear will remain a major obstacle.

While HCWs accept that to err is human, I don’t think the majority of HCWs ave an issue with commitment to improvement. It’s just that the fun of disclosure stops when litigation starts.

Jacqui (RSA)

Paul Levy said...

Thanks, Jacqui, but your final point is highly questionable. There is no evidence that this kind of disclosure leads to more litigation or more effective litigation. The kind of data we publish would be discoverable by a litigant in a malpractice case in any regard.

Jim Conway said...

Paul, three comments:
First, Congratulations to you and your whole team on your impressive leadership in reducing central line infections. As a BIDMC patient, I am thrilled you are all on this journey. As a healthcare and IHI executive, I’m proud to be presenting with you on the topic of transparency at the American College of Healthcare Executives Annual Meeting next month (guess it’s time to work on the talk!).

Next, when will we learn that nothing about healthcare infections is a surprise to the patients and families we serve that e experience them. They’re talking about them. Why don’t we? Madge Kaplan and I spoke to a group of retirees this week in Cambridge. In this small room of 20, we met “an infection,” “a wrong procedure,” “multiple failures to communicate,” “medication errors”, and yes, people who had great care givers and organizations they loved. In my experience, patients and families know this stuff happens; they don’t think we do.
Finally, over time, reducing infection, reduces cost and improves profitability. I was teaching at HSPH this week and I showed the clinicians and others a slide from the team at Virginia Mason. It pointed to the dramatic positive impact that improved quality and safety can have on profitability. Some told me it was counterintuitive—you could have increased quality and reduced cost. Your team is showing is that infection is a cost of poor quality and by attending to reducing infection, you will reduce cost, improve profitability, improve patient experience, and improve the experience of staff.

Paul, everything about this story is a great story. Yes, it is also a hard story. Let’s get our numbers out there so over time we can demonstrate to everyone that we are getting better in service to our patients, families, and staff.

Best wishes, Jim Conway, SVP, IHI.

Anonymous said...

It seems that fear of litigation is evolving into patent excuse to protect status quo. Your outrage is warranted, and your leadership is forcing a hand. In spite of the diminishing energy for change by some, do know that there is anger at the inaction simmering elsewhere.

What is the difference? The maker of peanut butter knows that salmonella sometimes gets in the door, and while testing to follow regulations is necessary, those data are irrelevant to the way work is done. Public opinion says the owner of that plant is culpable. The father of a child who dies from salmonella says the plant - and its regulators - are culpable.

A hospital knows that it is a potential petridish of harm. Steps are taken, efforts are made, but the data aren't important enough to share broadly. Fires are put out here and there, and quiet efforts are made to tamp down local problems. But transparency, internally as well as externally, is avoided, so the data lose their power to drive dramatic improvement in the infection environment.

If I were prone to sue for preventable harm, it would be the organization that suppressed the data, not the one aggressively working to solve the problems. Do lawyers take note of this difference? Would 'fear of litigation' for lack of transparency work in favor of quality improvement?

Steven Spear said...

An incredibly powerful barrier to delivering far better care, to many more people, with less effort and less cost is the lack of outcomes based measurement. No indication of performance means it is impossible for payers and patients to make informed choice. This rewards underachievers who hide behind the ambiguity of institutional reputation and hinders those pioneers trying to manage care delivery more effectively and efficiently.

CMS calls things like this "never events" because the should never happen. They are avoidable and when they do occur, it is do to poor processes and systems, not random events or acts of god.

Paul and colleagues, keep up the good work. You'll not only better the delivery of care at BIDMC but will prove that transparency is necessary for everyone else.

Respectfully yours,
Steve Spear
Senior Lecturer MIT
Senior Fellow IHI

Maureen Bisognano said...

Again, I applaud your use of transparancy as a lever for change, and your aggregation of the totals is a key to motivation. Two other levers include the aggregate costs of these defects and the human costs. Rick Shannon MD worked at two hospitals in PA to systematically reduce infections over several years. He then added up the dollar losses and challenged the senior teams with this question "what could we do with $10 million?" He had a slide with a list liek "Hire 166 nurses; get all patients with HIV in Haiti on meds" and the like. Great conversation starter!
From the human side, see "Ginny's Story" on YouTube....
Best wishes and don't lose your outrage...
Maureen Bisognano

Anonymous said...

So is this world expert working with medical school deans to ensure such training becomes embedded in medical school curricula? This is the only way you are going to ultimately change culture and attitudes in the medical profession. Believe me, I was one of the ignorant ones, till I was "converted" by an anesthesiologist - then all the advances we made with our Performance Improvement Council were lost when he went to another institution......gotta change culture from the bottom up.


Anonymous said...

The problem is that these issues are much more complex. Its hard to say that a certain mortality is attributable to central line infections as sicker patients get these infections and the mortality of not having a central line at the appropriate time because it are removed too soon for fear of line infections is also unclear. Just to post rate of line infections is simple but unclear if lower line infections is necessarily a good thing. Because JCAHO or medicare has chosen to follow these things is the "scientific" reason they have been followed. No one is tracking incidence of pneumothoraxes or other complications for lines that are removed too soon and then have to be replaced. By the same measure posting overall in-hospital mortality would also be a misleading fiure.

Paul Levy said...

Please! If you think another metric or set of metrics is better, propose it. Don't use the excuse of some amount of imprecision for a failure to track problems and success. That is a recipe for failure to act.

Judy said...

From someone on the tip of the spear, the change has to start in education and training. The culture has been, you make a mistake you get the rod from your superior because they know they will get it from theirs etc, etc. I'm speaking from a nursing viewpoint but have also seen it on the medical side.
I think this is why the whole Journey to Zero (or whatever your institution calls it)scares me; the stakes are high, the pressure is on-will the data be clean?
To make real improvements in healthcare we must remove the punitive response to reporting the truth.

Paul Levy said...

Right, a blame-free environment is a prerequisite to real success.

Carl said...


First off, I look forward to hearing you speack at ACHE's Congress.

Second,I am very impressed by your low rate of central line infections at your hospital. I do wonder, however, if this topic has anything to do with staffing resources. I have worked in a hospital setting for many years and have seen staff, very competent staff perform procedures quickly and sometime maybe not as careful as they should have. This was only because they needed to care for other patients, chart, etc. With healthcare cost above skyrocketing already and no end in site, at least for the near future, is this an unrealistic expectation of hospitals? Are we putting to much work load on great staff that we are causing burn out, shortages of nureses, doctors, tech, etc. Could this all be related in a big picture way of looking at healthcare today?

OMD said...

I am just doubting this metric in particular. How about disease specific mortality - like that for sepsis corrected for apache?

Anonymous said...

Accelerated movement in quality improvement at BIDMC includes full support by the governance culture. The Board strongly supports an environment where attention is driven toward positive variation, where innovation is rewarded, and challenges are on the table to be discussed collaboratively. Trust is driven by transparency, and visa versa.

Tom Dooley said...

I'm glad to see such spirited and open discussion. As a vascular surgeon at a large community hospital which refers many patients to BIDMC, I appreciate Mr. Levy's proactive style.

Paul Levy said...


I think this has less to do with the number of staff than with how work is organized among the staff you have. See Steve Spear's comment above and other things he has written.


Dunno. Beyond my knowledge. Others can comment.


Well stated!

Eva said...

Paul, congratulations on the great progress BIDMC has made. If you are looking for allies in quality improvement and patient safety, what about students? Students who have not yet been indoctrinated in the current health care culture are just as outraged as you are by inaction and failure. So many of these previous comments have referred to resistance in the current culture. Well, here's your chance to transform the culture. Teach us about how to function in a culture of transparency. Teach us how to share results of progress and success across institutions. Teach us how to prevent avoidable deaths. We are listening and are ready to take action. How can we help?

- Eva
future med student, IHI, Special Assistant to CEO

Bernard said...

Children's Hospital is Boston does show information about CLI, albeit in a roundabout way.

I was waiting in a pre-surgery area with my son last week and noticed various Windows laptops displaying a series of pages. I went to look at them and one said something like "106 days and counting since last Central Line Infection. Congratulations to (some department)."

I think it's important to make this information as visible as possible. In software we've found displaying measured values (bug counts, for example) generally causes the value to improve.

Michael said...

Google is promoting the idea that making home energy usage easily visible leads to improvements. It's a compelling argument, just as with software bug counts and hospital infection rates. What's needed is a commitment to improvement that outweighs complacency, inertia, and fear.

Anonymous said...

" . . . I was "converted" by an anesthesiologist - then all the advances we made with our Performance Improvement Council were lost when he went to another institution......gotta change culture from the bottom up. . . ."

It is hard to change culture if the people on the front line don't buy into the process change. It is sometimes necessary to give them the option to "Buy in or bye out".

the zak said...

Local dental schools' clinics need to be looked at with respect to infection control...Tufts University School of Dental Medicine Clinics Boston University School of Dental Medicine Clinics Harvard University School of Dental Medicine Clinics Forsyth Dental Clinics

Julie said...

Your indefatigable efforts to promote transparency are admired by many, including us at the Kenneth B. Schwartz Center, where we know that opening up the lines of communication with patients and their loved ones is ultimately good for everyone.

Studies show that effective patient-caregiver communication is
associated with stronger relationships, enhanced patient satisfaction, and reduced malpractice claims. Patients want their caregivers to talk
to them honestly, frankly and clearly. And they want their
institutions to do the same. Keep pounding away at that wall of
recalcitrance Paul. The wall will eventually fall.

Julie Rosen
Executive Director
The Kenneth B. Schwartz Center

Jon said...

I think you are correct, in that almost all of the quality indicator measures that are currently in place have problems. Some have poor evidence of benefits in mortality, some lead to unintended consequences (over use of antibiotics, or overly aggressive rush to Cath lab for possible MIs). I think it is important to thoroughly review the evidence behind these quality indicators.
Speaking of... the National Quality Measure Clearinghouse ( publishes overviews of thousands of different suggested quality measures. Some of these do list the evidence behind the measure, though it is rarely randomized controlled trial data. I recently reviewed a number of these indicators as they relate to emergency medicine, and the evidence behind them. It was eye opening. I think the key is to use a package of multiple quality indicators which are difficulty to "game" and to carefully watch out for the unintended consequences that these measures may lead to.

A number of organizations support the use of central Line infection rates as a quality measure, including our friends the Joint Commision. The Veterans Health Administration also use this measure. I've stolen part of their description for the evidence in favor:
"The CDC's National Healthcare Safety Network (NHSN) reported in July 2007 the median rate of CLAB infections per 1,000 central line days in NHSN participating hospitals ranged from 1.9 in Major Teaching Medical Surgical ICUs to 2.2 in Medical ICUs. The attributed mortality for these CLAB infections is mixed in the literature. The documented studies range from no increase in mortality when controlled for severity of illness, to a 35% increase in mortality in prospective studies that did not control for severity of illness. Further research to clarify the mortality associated with CLAB infections is needed, but the available data are consistent with an attributable mortality of CLAB infections, ranging between 4% and 20%. These infections are associated with 2,400 to 20,000 deaths per year."*

So, the data isn't perfect, but it trends towards lowering death rate. The second question is are we harming patients in a way these studies don't capture? For example, are we taking central lines out of patients too early? Is the full package of sterile techniques harmful in some way? I don't think so, but it deserves consideration.

On the topic of education, I put in two of these lines at work tonight, and was happy to teach one of my more junior colleagues the full sterile technique. It probably only took an additional minute to get the patient fully draped and the physician in a sterile gown. It took about 40 seconds to fill out the proper paperwork out to document our technique. In this particular case, the time/effort expenditure is certainly worth it, even in a busy practice environment.

*Full document @

Damjan DeNoble said...

My partner (Tucson Dunn, a former hospital CEO) and I were just discussing transparency issues in hospital billing system and how this is making American hospitals less competitive internationally when going head to head like Apollo Hospital Group facilities in India, or Bumungrad International in Thailand (to just mention two of the better known examples) which make publicly available their billing codexes.

The key point that Tucson made is that when a vigilant patient finds a suprise on their hospital bill (like a 1000 dollar toothbrush) it is tremendously difficult to get such a mistake (intentional or not) refunded, because, as you well know, the bill has to make it up to management level in the hospital. And this doesn't usually happen until 90 days have passed from the initial billing (that's when the clerks office sends the bill upstairs), at which point, the patient, through no mistake of his, sees his/her credit rating drop (the clerks office has been alerting the credit company at regular intervals for three months about not having yet recieved payment on a 1000 dollar bill and the credit company could care less if it was a toothbrush or an engagement ring) which point patients will take their complaints public and litigation and/or public embarassment for the hospital ensues. Meanwhile, all this could have been avoided if the bill was written so that a lay person could understand it, instead of being an endless list of letters and numbers worthy of a Dan Brown best seller.

This logic holds for the transparency issue you talk about. And, while I understand where much of hospital administrator's fears about transparency comes from at a time in our history when even good samaritans can get sued by those they save (see December 18 California Supreme Court Ruling allowing good samaritans to be sued for non-medical care) it is amazing to me that in this age of constant information flow, and of ever more prominent Health 2.0 philosophies perforating academic discussion, how hospital administrators can still be turning a deaf ear to the pleads of consumers for greater transparency. Mr. Levy, you are absolutely right about it being a life or death choice.


p.s. - long time reader, would love to link blogs -

Anonymous said...

Anon 8:24;

Well, in this case the people that didn't "buy in" were the new administration, so the converted were the ones to eventually "bye out". (Other members of the medical staff were essentially passive.) Nonbelievers are not limited to members of the medical profession by any means. My bottom up idea was predicated on the notion that if every doctor in a hospital is trained in the tenets of patient safety, it would become necessary and normal practice, not an evangelist's dream.

nonlocal MD

J Clevenger said...


The current situation will change, it has to. I think the entire health care industry is headed for serious change, which will hopefully benefit all of us.

Regarding the idea of implementing QC for hospitals, to reduce infections and provide better care, I think the only serious way to bring about this change is to emulate the QC processes of other industries. There should be an entire department at every large provider to deal with QC issues, and the QC management should control all operations. Procurement, surgery, clinics, engineering, etc. should all have to answer to QC. Annual meetings with the heads of all depts would be necessary, at which time revisions to the current processes would be proposed.

I'm sure you are familiar with this sort of operations structure, but I'm not sure if it is implemented at all in the hospital setting.

Paul Levy said...

I think most people will tell you that such an approach is at variance with how hospitals function, i.e., with MDs and others insisting on retaining control of the relationship with the patients.

Anonymous said...

My 84 year old mother, and many like her, refuse to go to the hospital, in fact live in deathly fear of going to the hospital becuase she and they know that once in, they are likely to die of infections contracted while there, if not for whatever acute condition brought them there. If BIDMC can succeed in publicizing it success at infection control, it would go a long way to easing the anxiety of potential patients who need early treatment but currently refuse it to their detriment.

Lonnie Fuller said...

We physicians are trained as master craftsmen, not as members of a team. As students, we go on attending rounds and try to outperform our peers with our exhibitions of knowledge. As residents in training, we learn that we have the power to speak and it will happen, that is, the intern will write the order, the unit clerk will take it off, and other folks will follow the order. The only teaching about how to deliver care was, "Stay until it's done." We go into practice trapped in a cultural crevice where we know more than any of the support staff, we do everything, we make all the decisions, we don't trust anyone else to do the right thing, even our peers and we live under the tremendous pressure that comes with trying to accomplish an impossible task given the methods we are trained with.

Given all that, I am still optimistic. The level of pain has risen to the point where even change looks good to some. You are part of that group. If you then follow Everett Rogers work on the Diffusion of Innovations, the next step is to make the benefits of the early adopters visible to others to influence the next wave of adopters. I do this in the rural counties of PA, where early adopters are geographically distant and functionally isolated. They light up and glow when they learn about others who are making the same change to empowered teams. They feel validated and they are willing to share and learn from each other.

Rather than challenging all the leaders in Boston to follow your lead, you might search out those who are ready to change and let the rest stay the way they are for now. Since only 13% of any population are early adopters, you won't get them all at once. However, driving along the adoption curve will bring them along and you won't have to wait for incremental change.

I am very glad I found your blog. Thanks for making your own efforts visible.

bill said...

Thanks for the openness Paul-

Has your hosptial investigated the use of nanotechnology in infection control?

How about the use of ATP monitoring to quantify cleanliness?


Paul Levy said...