Making Hospital Discharges Safer

Bob Wachter blogged about an uninteresting topic in an interesting way. The topic (medical errors made because of care transition "fumbles") is "uninteresting" because directing your patient's care elsewhere represents "closure". Who wants to dwell on the past? 'He's not my patient anymore. He's not my problem anymore.'

More on this critical issue here.Bob Wachter blogged about an uninteresting topic in an interesting way. The topic (medical errors made because of care transition "fumbles") is "uninteresting" because directing your patient's care elsewhere represents "closure". Who wants to dwell on the past? 'He's not my patient anymore. He's not my problem anymore.'

Of course this is the wrong attitude but call me a liar for proclaiming it a norm throughout our health care system. There was once a time when doctors coveted their patients like greedy misers. Woe to the doctor who stole a patient from his colleague! Woe to the specialist who usurped the PCP's hegemony! The psychology was different when fee-for-service medicine was the standard. With declining reimbursements caused by the rise of government and private third party payers as well as capitated medicine; this all changed years ago.

Medical residents weren't that covetous even back then but hospital care by the housestaff has always been the antithesis of fee-for-service medicine. Residents are paid the same regardless of their census. This led to the popular technique of "turfing" (transferring a patient to someone else's service). If these turf wars in reverse sound like a game, read the medical field's version of a cult novel, The House of God to get the full flavor.

Unfortunately, the health care system as a whole has adopted the medical resident "model". If the modern transfer is uninteresting, it's because the patient has just become someone else's problem.

With fractionation of care and lack of ownership comes patient dissatisfaction and worse, medical errors. Of course there are other reasons for the increase in disjointed care: the increased complexity of medicine, the greater acuity of hospital patients, the declining reliance on face-to-face or at least telephone communications in medicine. But, I think that this lack of ownership is way up there.

Wachter points out that no one is currently accountable for transition-related goofs. I think this is partially right. I don't think it explains everything. He points out the current lack of quality measures specifically related to determinining the integrity of the hand-off process. But I think we've always been aware of the problem. I believe that high readmission rates (which are measured) are generally attributed to poor post-discharge planning as a first approximation. Obviously there are other causes but this is usually what is blamed first. The principle culprit of this poor planning is frequently miscommunication.

He is right though that easily measured indicators that are more specific to the quality of hand-off procedures are required.

His discussion of "transition coaches" and his interview with Eric Coleman is fascinating. I'm quite surprised to read that such coaches were found to be cost-effective though. They seem to be extremely hands on, spending much time with patients, visiting them at home, etc. This seems very expensive to me. Clearly the patients eligible for this type of intervention need to be closely screened. I'd think that for such a program to work, only patients classified as being at very high risk of bad outcomes should be considered.

I once worked as a hospitalist for a large IPA that achieved good results with some relatively simple and inexpensive procedures. We immediately dictated detailed discharge summaries at the time of discharge. The IPA had a special contractural arrangement with the hospital to have these transcribed immediately and they were FAX'ed to the PCP (and relevant subspecialists) by the following morning. Our IPA discharge summaries had a standard format that was designed to be specifically helpful to the doctors participating in after-care (and that met the discharge summary standards for the hospital as well).

We also arranged for a full-time RN who called every single discharged patient the next day. She was often able to troubleshoot problems before they became problems. Of course one of the big advantages of such a nurse was that screw-ups became less serious and were more easily rectified. The transition coaches discussed above appear mainly geared at preventing screw-ups in the first place -- always a better though I believe costly solution.

There are programs being studied that assign roving physicians to do housecalls on frequent flyer patients (often old, tenuous patients with multiple medical problems). Such programs appear to be cost-effective but are treating a somewhat different problem i.e. trying to prevent admissions in the first place.

One thing that really used to disturb me was the response of PCP's to phonecalls. If I was discharging a particularly challenging patient, I would call the PCP to give him a heads up. Surprisingly, the most common response was "Don't bother. Just FAX the D/C summary." They didn't even want to take the two minutes to discuss it.

Now if there's just a way of financially incentivizing doctors to take ownership (in ways that don't lead to overutilization) then we'd really have something!

Everyone be sure and have a great new year!

Rapid Response Teams: Still Wondering After Allllllllll These Years

Bob Wachter blogged about Rapid Response Teams (RRT's) today. His observations about this often misunderstood hospital safety intervention are quite helpful.

Click here for my thoughts on this.Bob Wachter blogged about Rapid Response Teams (RRT's) today. His observations about this often misunderstood hospital safety intervention are quite helpful.

RRT's are teams of personnel that can be summoned when a hospital patient isn't critically ill but is starting to just look bad. Dr. Wachter describes them as a Code Blue team in drag. I choose to think of them as a Code Blue team minus the adrenalin. Nurses are more willing to call them than call a Code Blue team because they're smaller and just...not as big a deal.

He is quite right that the empirical evidence supporting RRT's is lacking. Some relatively small studies (see here and here) did show benefit in terms of reduced mortality but these were not well "controlled" and their results can be easily refuted. The one large, multi-center trial, the Australian MERIT trial, failed to demonstrate benefit.

As Dr Wachter reports, the Institute of Healthcare Improvement (IHI) was so wedded to the RRT concept, that they elected to promote it as a national standard. Clearly, they were disturbed by the less then stellar performances RRT's exhibited in the MERIT trial and felt compelled to publish a specific critique of that study.

This critique struck me as quite reasonable. They noted that both the control and RRT groups demonstrated a reduction in mortality during the study period. IHI suggested that this could be explained by noting that in Australia, RRT's were already being adopted in various incarnations throughout the country. This probably tended to make the control and study groups look more alike and therefore lowered the likelihood of demonstrating a difference.

The IHI also made the point that the actual implementation of the RRT's in the study groups may have been suboptimal (and gave examples of this). Had the study groups RRT's been better executed, again, true benefits may well have been realized.

Arguments were also made that the hospitals studied were largely academic institutions and as such, the results may not have been generalizable. Likewise, they raised arguments that the study as a whole may have been too small and may not have had the statistical power to show a true difference.

For myself, I tend to think that there is some benefit to these programs even if the proof is hard to arrive at. The key, I think, is appropriate implementation.

I've seen situations where the RRT begins to look very much like a Code Blue team. This doesn't strike me as cost-effective. Too many people, too many doctors, too big a deal. I much prefer simpler concepts such as the version that Dr. Wachter eventually organized in his institution which seems to mainly be run by ICU nurses and respiratory technicians used to assessing very ill patients.

That such teams can be helpful in the event of a "pre-emergency" makes sense to me. In the second study I cited above, there is a very interesting table that documents the types of interventions the RRT's used. The leaders of the hit parade were:

1. Nasopharyngeal suctioning with additional oxygen
2. A simple IV fluid bolus
3. A dose of IV Frusemide (the diuretic Lasix for those of us on this side of the Atlantic)
4. Noninvasive positive pressure ventilation by mask
5. Nebulizer treatments
6. Temporary manual ventilation
7. Trach tub suctioning
8. IV Glyceryl trinitrate drip (which I had to look up to convince myself was just good ol' nitroglycerin)

With the exception of 8. and possibly 4. (which has some subtleties to it that can cause damage), these are all simple and relatively harmless interventions whose decision to initiate can easily be made by an experienced nurse. We're not talking major stuff here; just some simple things that can make all the difference in the world to a patient that's not quite critical but whose condition is worsening.

It's hard to argue with the utility of such an approach. I agree with Dr. Wachter that (at least given what we "know" at this time), individual institutions should be allowed to decide for themselves and later assess how well they work and whether they should be continued.

One additional point. I've heard a number of hospital nursing directors say that RRT's are a very useful nursing recruitment tool. It seems quite clear that nurses are very happy with RRT's and probably see them as helping to alleviate the anxiety of taking care of potentially deteriorating patients.

That has to be good!

Electronic Alerts to Improve Hospital Outcomes

Here is the reference to an article that appeared in the New England Journal of Medicine. Basically it describes a very innovative randomized study that evaluated an "electronic alert" system. These alerts were computer-generated to notify admitting physicians that inpatients of theirs meeting specific criteria were at high risk of thromboembolic events (e.g., blood clots in the legs or lungs).

Half of the patients were subject to the alerts and half were not (the standard of care). These alerts reminded the admitting physician that it might be advantageous to place such patients on some type of prophylaxis such as anticoagulants, compression stockings, etc. The physician could then elect to apply such prophylaxis or not depending on his preference. Hospitalized patients can be at risk for such events for a variety of reasons and the amount of morbidity and death is significant.

It turned out that physicians taking care of the patients subject to the alerts were more likely to place their patients on prophylaxis then with the control group. More importantly, those patients did better and were 41% less likely to actually have an actual adverse event then the control group.

So this relatively simple, easily engineered computer program had a marked impact on patient outcomes.

As a hospitalist (one who specializes in taking care of inpatients in acute care facilities) attempts by hospitals to initiate such interventions have always been annoying to me. While admitting patients with certain medical problems, I have often been confronted with pre-printed admission orders for given diseases used to gently prod me to address specific issues.

I had always found such pre-printed orders annoying. I felt that I as an experienced physician, I knew how to take care of my patients and the last thing I needed was the product of some bureaucratically entangled "forms committee" to tell me how to manage patients. Nor was my (occasionally pig-headed view?) rare among doctors.

Studies like this with such dramatic results are very hard to ignore. There is no doubt that as physicians involved in patient care, we are going to see a lot more of this type of intervention. It may be annoying (and perhaps personally embarrassing to admit to ourselves that we missed something) but after all, it's the patient outcome that matters.

Measures such as this can make hospitals safer places to be.

(I violated my usual rule of never updating my posting because I wrote a really stupid sentence construction!)