Wednesday, September 27, 2006

HIV and the ostrich

In a recent speech I had the opportunity to attend, Dr. Julie Gerberding, director of the CDC, illustrated a point about meeting the challenges of the future by bringing up the ostrich’s supposed behavior of sticking its head in the ground at the first sight of danger. She cited the relevant Wikipedia article, which reads:

. . .there have been no recorded observations of this behavior. A common counter-argument is that a species that displayed this behavior would not likely survive very long. . .burying their heads in sand will in fact suffocate the ostrich. When threatened, ostriches run away, but they can also seriously injure with kicks from their powerful legs.

I like the way that this logic applies to HIV testing and the new U.S. federal policy of recommending annual testing for at-risk teenagers and adults. Living with HIV, but refusing to be let aware of it and thereby foregoing the possibility of life sustaining treatment, like the ostrich burying its head in the sand, is a behavior incompatible with life. It also poses a substantial threat to public health. The current policy climate in most states of validating such behavior by requiring documented consent before testing simply no longer makes sense.

Consenting somebody for HIV testing is more nuanced than it sounds; you don’t just ask the patient whether or not you may test him or her. Obtaining informed consent means explaining the pros and cons of the proposed action, and for HIV, I have tended to say something like the following:

Sir, before testing you for HIV, as we are suggesting, I have to obtain your written permission. The reason for this is to make sure that nobody is tested without their consent or knowledge. The benefit of testing is that then we will actually know whether you have HIV, and if you do, you will be able to receive life saving treatment. The downside of testing is that if you do have HIV, it can be difficult news to take, and in the past there have been certain social stigma associated with the disease. Of course, I strongly recommend testing.

A speech like this immediately raises suspicions--if testing is such a good idea, why does it require documented consent? Clearly, HIV is not an easy diagnosis to receive, but neither are a host of other chronic diseases from herpes to leukemia. Can you imagine your physician asking you for documented informed consent before testing you for say, schizophrenia? I’m not suggesting that HIV testing should be mandatory; within reason physicians should obtain permission before testing for any disease, but let’s stop this charade of pretending that declining or delaying testing is a reasonable choice. In DC, where one out of twenty of our metro riders carry the virus, annual HIV testing for at-risk teenagers and adults makes particular sense.

Saturday, September 23, 2006


This word gets thrown around way too much.

Work that you cannot begin to imagine achieving yourself, that displays a ruthless, nearly inhuman element of self-enclosed perfection—this is his idea of genius.
--Ian McEwan, in his novel Saturday,
which chronicles a day in the life of neurosurgeon Henry Perowne.

Thursday, September 21, 2006

Two diagnoses is the gracious approach

One diagnosis is the ideal of medicine, the most elegant and parsimonious way of explaining patient symptoms and findings. The patient with shortness of breath, pallor, weight loss, and chest pain, for example is not an anorexic asthmatic who snorted cocaine, but a victim of colon cancer, whose disease by chronic occult bleeding caused severe anemia and demand cardiac ischemia. This concept, one of medicine’s most revered principles, is taught to us early in medical education and hounded into our minds thereafter.

Two diagnoses, however, as an attending recently noted on rounds, is the more gracious approach. When a senior team member (i.e. attending) suggests an unusual diagnosis to the team when a more common diagnosis is already known to be active, the tactful resident or fellow will offer that both diagnoses may be contributing. (“Well, we thought the anemia was probably due to her severe iron deficiency, but certainly she could be hemolyzing as well.”)

Put two distinguished attendings on rounds together and you’ll never see so many dual diagnoses, our attending commented.

Of course, if you open your mouth as a medical student, you’ll be shot down before you even know what happened. A favorite teaching tactic of attendings, in fact, is to make you tell them why the diagnosis you just suggested is unlikely (or very unlikely). To prevent such humiliation, most students and residents will preface any proposed diagnosis with one or two reasons why it doesn’t quite fit. Such are the growing pains of medical education, and appropriately so; after all, one diagnosis is the ideal model for medicine. Nonetheless, we can take comfort in the knowledge that some day, when we have reached the necessary level of seniority and respect, we too may be indulged with that second diagnosis.

Monday, September 18, 2006

Top ten most dreaded night float pages

Here's a list that I compiled last year while covering intern night float, which at our institution means cross-covering on 100+ patients for all the major and minor problems that occur after hours, or as one of our attendings puts it, the "things that go bump in the night."

Most of these contain inside humor, usually the implication of an especially onerous task. I left out "Doctor, I wanted to let you know that your patient's blood pressure is 135/70. . .oh really? You're not concerned?. . . you just had fallen asleep?. . . sorry, doctor" as it fits more in the annoying category than dreaded one.

10. Doctor, Ms. Adams is crying again, could you please go talk to her?
9. Doctor, the patient was just found down on the floor. . .no, we’re not sure how long he’s been there. . . no, he doesn’t know how he ended up there. . .
8. Doctor, the lab just called about Ms. Baker, you know, the one you were just examining. . .she has scabies. . . you mean you weren’t wearing gloves?
7. Doctor, I just finished vital signs rounds. . . you may want to sit down for this. . .
6. Doctor, the patient is refusing to take her psych medications. . . she says the voices tell her she shouldn’t. . .
5. Doctor, the patient’s iv just infiltrated. . .but doctor, he has no veins!. . .well doctor, his iv antibiotics were supposed to start 4 hours ago and his temperature is 102.4. . .
4. Doctor, the patient still hasn’t had a bowel movement and is complaining of worsened bloating and abdominal pain. . .no, that’s right, neither the fleets, soapsuds nor oxbile enemas had any effect. . .
3. Doctor, the patient’s pain seems to be improved, but she is no longer waking up. .
2. Doctor, the patient just pulled his NG tube out again. . . yes, he’s still wearing the four point hard leather restraints and straitjacket. . .
1. Doctor, the patient’s pulse just shot up to 230, his blood pressure is 240/120, and he’s complaining of chest pain. . . no, only two of the three bags of cocaine that he swallowed have passed so far with the golytely

Saturday, September 16, 2006

HIV on the metro

A rather chilling poster decorated the insides of our metro trains this spring, carrying the following message:

For every 20 people aboard this train, one is infected with HIV.

That's the language I remember, anyway; I can’t currently seem to find a copy of it. The poster's purpose being to shock and scare, that’s exactly what it did. I remember one evening riding home from work, seeing the poster, and thinking that there must have been several hundred people on the train, perhaps a few dozen of whom carried the virus.

Of course, it obviously wasn’t true. The author would have used DC HIV infection rates, ignoring that most metro riders live outside of the city. But just how close to the truth was it? From the never-ending stream of HIV infected patients at the different hospitals where I work, I knew there must have been some basis for the claim, but do we really have an HIV epidemic in DC of African proportions?

According to the DC Department of Health,

As of December 31, 2004, the reported prevalence of HIV (not AIDS) in the District of Columbia was 17,205 cases. Of that total, 81.7% was African American, 14.6% was White,3.2% was Hispanic, and less than 1% was Asian/Pacific Islander and American Indian, whereas 70.1% was male, and 29.9% was female.

Considering there were 572,059 people living in DC, that makes a prevalence of 3.01%. The Department of Health estimates that up to 25,000 people (4.4%) may actually carry the disease. Round that up to 5% and there’s your poster.

How does DC compare to other states? Not very well. For persons living with AIDS, DC leads all other states with 1.906 per 100 people, over 4 times higher than New York, the next closest state. To be fair though, let's look at how Washington compares to other cities. These are the 5 metropolitan areas with the highest yearly AIDS incidence (newly diagnosed disease) per 100,000 people (2004 numbers):

1. New York 56.7
2. Washington DC 40.3
3. San Francisco 33.5
4. Orlando 31.2
5. Tampa 25.7

DC’s got HIV, and it's got it bad. Even if posters like the one on the metro exaggerate the threat, I’m glad to see someone’s doing something about it.

Tuesday, September 12, 2006

The lovenox story, part II

The lovenox story part I

Removed for the time being because I was unhappy with how the story came out. If you have an interest in this particular story, let me know and I can e-mail you a copy.

Monday, September 11, 2006

The lovenox story, part I

Removed for the time being because I was unhappy with how the story came out. If you have an interest in this particular story, let me know and I can e-mail you a copy.

The lovenox story part II

Thursday, September 07, 2006

Our patient?

For as long as I can remember, I’ve been introducing patients in my presentations on rounds as “our patient.” For example, “Mr. X is our 54 year old male with HIV and pneumonia.” Most of the other residents and students I’ve noticed do the same. Well, earlier this week our attending, who trained in the Age of Giants, suddenly and impulsively asked in the middle of a student presentation, “When did it become ‘our’ patient?”

We had all kinds of answers for him: “it IS our patient”. . .”it’s natural”. . . “it’s taking ownership of the patient”. . .”it’s just nicer.”

He harrumphed and didn’t really reply. During the next presentation, however, when a resident presented “our patient,” he was quickly corrected to “the patient.” After 3-4 more corrections during the rest of rounds, our attending effectively purged the words “our patient” from our vocabularies. I resented this at first; why on earth, I thought, was he doing this? After listening to several presentations with the new phrasing, though, I began to agree with the attending. “Our patient,” although certainly neither incorrect nor inappropriate, did carry a sort of pseudo-intimacy that colored the clinical picture. It just wasn’t really necessary and perhaps sounded a little too cutesy for academic medicine. Unencumbered by this false emotional attachment, “the patient” was just the cleanest and most simple way of presenting a patient.

Amazing, the feeling when you’ve been unwittingly looking through a rose-tinged lens for as long as you can remember and then somebody suddenly removes it.

Monday, September 04, 2006

Dining companions

This past weekend I stooped to a new low while on call. It had been a relatively slow Saturday, but I’d had three admissions interspersed over the afternoon and evening, and by the time I finished admitting the last patient, a high school-aged boy named Tim, an empty gnawing sensation had somehow crept its way into my stomach. When I went back into Tim’s room to follow up on a question I had forgotten to ask him earlier, I was surprised and jealous to see and smell that he had ordered Chinese food from a neighborhood restaurant. Now in my defense, the cafeteria was closed by then, and the only food available was the snacks in the vending machines. As I went over the details that I had returned for, I couldn’t help but take a few looks at the impressive amount of food that was sitting around his room, much of what appeared to be leftovers. “How was the food?” I asked him, “That looks really good.” We chatted a bit more, and I snuck a few more looks at the apparently left over chunks of deep-fried sweet and sour chicken. Finally the offer came, “I’m not going to eat all this food, would you like some? I promise I didn’t spit in it or do anything weird.”
“Well, it does look pretty good,” I said, offering no protest. “Sure, I’ll have some!”
I sampled one of the chicken pieces. “Not bad.”
“Take the whole carton,” he said.
“You’re sure you aren’t going to eat any more of it?”

I had just scored a free dinner from my teenage patient. I have to admit, I felt a little dirty about the whole thing, and to boot, I didn’t have a clue what germs might have landed on that chicken. My stomach had no qualms, though; it was feeling a whole lot better already. Somehow, I sensed that Tim felt good about the interaction too. Here was a kid who had been sick all of his life with a devastating congenital disorder predisposing him to all kinds of different infections. He had always been different, and he looked the part; his disease had stunted his growth and badly damaged his hair and skin. Yet, there I was, a young male medical resident, casually accepting his offer of food, demonstrating not only that I didn’t consider him unsanitary or covered in cooties, but that I also trusted him and thought that he was cool enough to eat with. Giving is an obvious way of demonstrating friendship, but sometimes being willing to put yourself into somebody else’s debt says just as much. In admittedly odd fashion, we reached out to each other, and in doing so we set a positive and trusting tone to the rest of our interactions for his hospital stay.

Saturday, September 02, 2006

Double coverage and pseudomonas, part II

Having spent some time looking into this matter and benefited from some ID tutelage, I can affirm that there are two general justifications for double covering pseudomonas and/or other gram negatives. The first is to ensure appropriate coverage; with the high incidence of multi-drug resistance bacteria in our hospitals currently, we want to make sure that our monotherapy doesn’t consist of a drug to which the bacteria is resistant. The second justification, as the pharmacist suggested, is indeed to prevent the emergence of further antibiotic resistance. Let’s evaluate each of these rationales.

But first, let’s familiarize ourselves with the profile of the enemy. Using pseudomonas as our poster child, antibiotic resistance patterns measured in the DC VA in 2004 were as follows:

Anitbiotic Percent susceptible
Piperacillin 78
Ceftazidime 79
Gentamicin 83
Tobramycin 88
Aztreonam 64
Ciprofloxacin 63
Levofloxacin 68
Meropenem 89

Improving outcomes
So, does double covering severe gram negative infections reduce mortality or lead to better patient outcomes? You would think so, considering that for the single antibiotics we would use there’s a 10-20 percent chance that the bug is going to be resistant. The evidence, though, is less than impressive.

The main support for double covering pseudomonas seems to come from a meta-analysis by Safdar et al. published in the Lancet in 2004. Their meta-analysis included 17 studies looking at monotherapy versus combination therapy for patients with gram negative bacteremia. Most of the studies used beta-lactams and aminoglycosides in single and combination therapy. While overall there were no differences in mortality, analysis of the Pseudomonas bacteremias showed a quite substantial mortality benefit (OR 0.50, 95% CI 0.30-0.79) with combination therapy. The authors concluded that combination antimicrobial therapy was indicated when there was a strong suspicion for Pseudomonas, but not otherwise.

Contradicting Safdar’s findings, Paul et al. conducted a large meta-analysis of 64 trials including 7586 patients comparing beta-lactam monotherapy versus beta-lactam-aminoglycoside combination therapy for sepsis. They found that double coverage conferred no advantage over single coverage in sepsis, even when gram negative and Pseudomonas infections were analyzed separately. They did find more nephrotoxicity in the double coverage group, and they concluded that aminoglycosides should not be added to beta-lactam therapy in sepsis.

They also did not find that double coverage affected the development of resistance.

Other studies:
De Pauw et al.: Ceftazidine was safer and as effective as the combination of piperacillin and tobramycin for empirically treating febrile neutropenic patients.

Eggimann and Revelly: ventilator associated pneumonia should not be treated with antibiotic combination therapy, which may promote the emergence of drug resistant bacteria. Cunha and Damas et al. draw similar conclusions.

From this, I would say that double covering for pseudomonas makes sense for patients with unspeciated bacteremia where pseudomonas is thought likely, but not in other circumstances.

Preventing Resistance
As far as preventing antibiotic resistance, there seems to be little evidence. The Infectious Diseases Society of America guidelines lists the following factors that can contribute to antibiotic resistance:

Greater severity of illness of hospitalized patients
More severely immunocompromised patients
Newer devices and procedures in use
Increased introduction of resistant organisms from the community
Ineffective infection control and isolation practices and compliance
Increased use of antimicrobial prophylaxis
Increased empiric polymicrobial antimicrobial therapyHigh antimicrobial usage per geographic area per unit time

They make the following comment:

The last strategy suggested in table 6, use of combination antimicrobial therapy to reduce emergence of resistance, is theoretically attractive and is the basis for current treatment of tuberculosis with multiple antimicrobials. It has not been adequately tested clinically to determine if overall institutional resistance can be reduced by the use of combination therapy for individual patients [47]. In one study of Enterobacter, no benefit in reducing emerging resistance was observed when combined third-generation cephalosporin and aminoglycoside therapy was used [48]. The risks include increased antimicrobial costs and the potential for increasing resistance by raising the number of antimicrobials and antimicrobial courses administered. The use of combination therapy is, however, already widespread for the treatment of seriously ill patients, so controlled trials to determine the effect on resistance prevention are reasonable.

It seems to me that the additional risk of toxicity from these powerful antibiotics probably outweighs the potential benefit of delaying the development of drug resistance, when there’s also a chance that the additional antibiotic exposure could actually speed up the development of drug resistance. I have a hard time justifying this one. Anybody with a different take?

Friday, September 01, 2006

Double coverage and pseudomonas, part I

Enough of this fluff, it’s time for a little meat. Here’s a clinical problem I experienced as an intern last year: the pharmacist called because my patient with a leg ulcer was growing out pseudomonas, and she was concerned that the patient was only on single coverage with zosyn.
“The pseudomonas is sensitive to zosyn, though, right?” I asked.
“Yes,” she confirmed.
“So why would we want to double cover?”
“Well, usually it’s recommended to double cover against pseudomonas so it won’t develop resistance.”
“Hmm, I’m not sure that makes sense to me – I thought we only double covered empirically until we found out the sensitivities – let me talk with my team and get back to you.”

So who was right? Discussion to follow. . .