An ongoing dialogue on HIV/AIDS, infectious diseases,
August 3rd, 2011
Why the Proposed Massachusetts HIV Testing Bill is Bad for Patients
As I’ve written about here multiple times, I’m not a big fan of the HIV testing law in our state.
First, there’s the requirement for written informed consent, something that every state (except a couple) has wisely abandoned. Second, it’s more than a testing law — it’s also an HIV privacy law, which is arguably unnecessary in this post HIPAA era and has all sorts of unintended consequences.
Last year, two bills were proposed — one simply removed the requirement for written consent, while the other replaced it with a requirement for verbal consent. The groups backing these respective bills (I was part of the former) didn’t work out a compromise, and so no bill was passed.
This year? Something very strange has happened. We have a bill pending “An Act to Increase routine screening for HIV” that would remove written consent but would also expand HIV-related privacy protections and mandate that primary care and ID providers “offer” HIV tests to their patients. The bill is working its way through government channels, on its way to being passed.
And you know what? I’m not aware of a single ID or HIV clinician who supports it.
This includes providers at hospitals, community-based clinics, group practices, and the largest medical practice in the state for gay and lesbian people. That’s right — most of the people dedicating their careers to HIV care don’t like it.
Here’s why this bill is a problem:
- It requires documentation that verbal consent for HIV testing was obtained. I’ve been told that in NY, where a similar law exists, many institutions haven’t removed the requirement for written consent because of this provision — patients still have to sign a form. In short, practice hasn’t changed even though the law has.
- The definition of protected “HIV-related information” has been broadly expanded, with a requirement for written patient consent before this information is shared. As currently written, the proposed law would block clinician-to-clinician referrals if they are not in the same facility, or a simple release of information from a hospital to a community PCP — discharge summaries, test results, diagnoses, medications. All would require written consent from the patient first. (Note that this goes well beyond the existing requirement for written consent to release medical records — this we already do for everyone, not just those with HIV.) Needless to say, no other disease has this restriction, and to say this is an obstacle for good patient care is a huge understatement. How would this even be applied to electronic medical records, which are increasingly being used in the community?
- There is language mandating that primary care and ID clinicians must offer an HIV test to a patient unless it is documented that the test has been done before. Is there another screening test (colonoscopy, mammography, TB skin testing, hepatitis C) where a law exists that clinicians must offer the test? Does this increase a clinician’s legal risks around HIV testing? And how do we access whether the test has been done if it’s protected information, as noted in #2? Recommendations for HIV testing should fall under the province of clinical guidelines (such as those issued by CDC), not under law.
- There is no way to test critically ill patients who are unable to give consent unless they have a health care proxy to consent for them. Anecdote: I saw a 35-year-old woman with a large brain abscess late last year who was comatose. An HIV test would have been invaluable to help distinguish bacterial brain abscess from possible toxoplasmosis — conditions which obviously require different therapies — but she couldn’t give consent and, as is very common in younger patients, she had no health care proxy. Wouldn’t it make more sense to allow HIV testing in instances like this, where the treating clinician could provide better care if he/she knew the patient’s HIV status? This is obviously how testing for other conditions in critically ill patients is done. Notably, the argument to allow testing for HIV in this setting was brilliantly made by a bioethicist writing way back in 2005.
- There is no way to allow testing of source patients in occupational exposures unless the source gives consent. In some states, antibody testing on previously-obtained blood samples is permitted. I’ve expressed this view before, but in occupational exposures, the benefits of knowing a source patient’s HIV status extend to both the source patient and the exposed health-care worker. Shouldn’t we do everything we can to make obtaining this information as easy as possible?
I know the proposers of the bill have the best interests of people living with HIV in mind, but the problems listed above suggest that clinicians were not part of the development process. Since the bill has been on the table, several HIV clinicians and researchers (including me) met with the sponsor to discuss our concerns, but to no avail. For the record, the bill is also opposed by the Massachusetts Medical Society, who articulated these concerns at a public hearing earlier this year.
So let’s change the law, yes. But let’s get it right please:
- Remove the requirement for written consent.
- Allow existing privacy protection statutes (which did not exist in 1986, when the current HIV testing law was enacted) to apply to HIV.
If you live in Massachusetts and care about this, find your Senator and Representative, and let them know that this bill is bad for patients — and has the unintended effect of being discriminatory, leading to inferior care for people living with HIV.
You can bet I’ll do the same.
July 28th, 2011
Really Rapid Review — IAS 2011 Rome
Just back from IAS 2011 (which was followed, I’m thrilled to say, with a visit to perhaps the most beautiful region in the world). Here is a Really Rapid Review™ of the meeting, with apologies ahead of time for lack of organization and (even more likely) leaving out something important. FYI, the abstracts are online here; I’m sure there’s a place on the meeting homepage that has the same link, but I can’t seem to find it.
- All 052, all the time. This landmark study was the big story at the conference, with several presentations on it (and the paper published the same day in the NEJM; Journal Watch comment here). Very interesting fact: the only linked transmission in a person whose partner was randomized to early therapy probably took place before virologic suppression. At the end of the series of talks on the study — which covered the primary outcome, details on how they documented viral linkage, and the clinical, immunologic, and virologic results — the investigators (led by Mike Cohen) received a standing ovation from the audience. Well deserved. (Quick aside: Did they ever tell the participants whether the HIV transmission was linked to the infected partner? I suspect not, but …)
- What about PrEP? Results from two favorable studies (abstract here for one, slides here for the other) were released just before the conference, and further details were presented right after the 052 data. Important studies? Yes. “Game changers” for clinical practice, like 052? Probably not.
- Dolutegravir looks really, really good. Week 48 data from a relatively large phase II naive study show around a 90% suppression rate, no obvious toxicity, and — amazingly — no integrase-inhibitor resistance among virologic failures. Could dolutegravir be like a boosted PI in that regard? This smart guy thinks so. In an odd coincidence of recent HIV drug development, dolutegravir seems to do the same thing with inhibition of creatinine secretion (thereby raising serum creatinine without reducing GFR) as cobicistat.
- Speaking of integrase inhibitors, once-daily elvitegravir is “non-inferior” to raltegravir in treatment-experienced patients. Good news, as the phase II study of elvitegravir was highly problematic, in hindsight due to both dosing issues and the absence of active background therapy. Note that in this phase III study, elvitegravir was given with a PI boosted by ritonavir, not by cobicistat. And cobicistat sounds great with an Italian accent.
- Any other promising new investigational drugs out there? Maybe –The NNRTI lersivirine was almost as good as efavirenz in this phase II study. Since phase II studies tend to be relatively small, any signals of inferior efficacy are necessarily a concern.
- In a similar vein, the two-drug regimen of ATV/r and QD maraviroc was (again) almost as good as ATV/r + TDF/FTC. Specifically: More hyperbilirubinemia in the maraviroc arm and also more low-level viremia. Since the population had to have a CD4 > 100 and R5 virus, the bar was not set very high for successful treatment. Could three drugs be required for optimal efficacy no matter what?
- Week 96 data on TDF/FTC + rilpivirine show pretty much what we saw at week 48 (published in this issue of The Lancet): More virologic failure with rilpivirine (especially at high viral loads), but more issues around tolerability/lipids with efavirenz. An open-label study of the two single-pill options using these drugs — Atripla (TDF/FTC/EFV) vs. so-called B-tripla (TDF/FTC/RPV) — is ongoing.
- What happens if you take 311 patients stable on ABC/3TC + a boosted PI and randomize half of them to change the NRTIs to TDF/FTC? Yes, lipids improve and estimated GFR worsens, but there’s also something surprising: fewer virologic failures (3 vs. 11). All low-level virologic rebounds, but still — perhaps the results are telling us something about the relative potency of these two NRTI combinations.
- We’ve known for some time that tenofovir-based treatments reduce bone-mineral density more than comparators, but what are the clinical sequelae? In this large VA-based study, tenofovir exposure was associated with a small but statistically significant risk of osteoporotic fractures (hazard ratio after controlling for everything, 1.12).
- Do boosted PIs increase the risk of chronic kidney disease? Perhaps some of them — notably ATV/r and LPV/r — do, according to these three studies (one, two — abstract says DRV/r, poster didn’t — and three). Is this mediated through their effect on tenofovir levels or some other mechanism?
- Treating acute HIV infection for 48 weeks delayed CD4 decline more than treating for only 12 weeks or not at all. Viral “set point” after stopping treatment was also lower in the 48-week treatment group, with a greater beneficial effect seen if therapy was started closer to seroconversion. Since this was a randomized trial, and since treatment interruption has looked pretty bad in every recent study, are we ready to say that all patients with acute HIV should be treated? I think we’re very close, if not there already.
- Does a regimen’s “CPE” (CNS penetration effectiveness) score improve neuropsychologic function? Not according to this study. Data in this area are so conflicting that one almost needs to be a top performer on neuropsych tests just to keep the studies straight.
- With d4T use being phased out worldwide, the four WHO-recommended initial regimens all consist of tenofovir, 3TC or FTC, and either efavirenz or nevirapine. According to this presentation, TDF/3TC/NVP could be suboptimal, with high rates of virologic failure and emergent drug resistance (especially K65R). Given that this regimen is probably the cheapest and thus the most likely to be implemented widely, confirming its effectiveness is absolutely critical.
A few final words about the conference: The weather was beautiful (one shudders to think what next summer in Washington DC will be like), Rome is extraordinary (so long as you can dodge all the scooters and Fiats zipping around), and I bet all 7,482 participants ate very, very well. Certainly I did.
But did the posters have to be displayed in a converted parking garage?
July 13th, 2011
More Favorable Results on PrEP, But …
As part of the usual flurry of studies released just before major scientific meetings, results of two pre-exposure prophylaxis (PrEP) trials in heterosexual men and women have just been made public:
- In the CDC TDF2 study, 1200 HIV-uninfected men and women in Botswana were randomized to take oral tenofovir/FTC or placebo daily. Tenofovir/FTC was found to reduce the risk for HIV acquisition by 63%. The risk reduction was even greater (78%) among individuals believed to be taking the study drugs.
- In the Partners PrEP Study, the uninfected partners in nearly 5000 HIV-serodiscordant couples in Kenya and Uganda were randomized to take oral tenofovir/FTC, oral tenofovir alone, or oral placebo daily. Relative to placebo, tenofovir/FTC was associated with a 73% reduction in risk for HIV acquisition, and tenofovir alone was associated with a 62% reduction in risk. The protective effects were similar for both men and women.
So, the score so far on PrEP studies: four positive (if you include CAPRISA 004, which used tenofovir vaginal gel, and of course iPrEX) and one negative (FEM-PREP, in women). Other key (if unsurprising) findings are that PrEP works better if you take it and that it’s pretty safe.
One possible interpretation of these results is that we should do everything we can to get HIV-negative individuals from serodiscordant couples on PrEP as soon as possible.
But here’s another thought: Why not get all the infected people — especially in established serodiscordant couples, as in Partners PrEP — on ART? According to Study 052 (with many more details to come next week at IAS), the net result of this strategy would be both a personal and a public health benefit. Sure, HIV treatment with three drugs is more expensive than one or two for PrEP, but it strikes me that the double benefit (i.e., to the individual and the uninfected partner) easily justifies the incremental cost.
If the final results of 052 are as impressive as the glimpse we’ve received thus far, one could see oral PrEP becoming a strategy limited to high-risk individuals who do not have steady partners (e.g., MSM similar to those in iPrEx, or certain individual from hyperendemic areas).
Which means that the best way to prevent HIV with meds is usually to treat those who have it — and only rarely to treat those who don’t.
July 12th, 2011
A Thank You to Nice Patients
Yesterday I saw one of my favorite long-term patients. She’s just wonderful, and it’s always a joy to see her. Here are some reasons:
- She’s uniformly nice — not just to me, but to all the nurses and social workers and other support staff in our clinic.
- She shows up on time for her appointments. Love that.
- She’s highly involved in her care, but doesn’t try to make all the decisions. Example: when I proposed a medication change a few years back, she asked many thoughtful questions about the pros and cons of the change I was recommending. But she didn’t then come back with reams of internet research questioning everything we’d already discussed. In other words, she gets it that I do this full time for a living, which means that she doesn’t have to do the same.
- Even though her kids call her all the time on her cell phone, she cuts the conversations short during our encounters. Then she politely apologizes for the interruption.
- She’s very appreciative. At least she seems to be — and it’s great to be appreciated! Some of our patients send us a card around the holidays. I’ve saved every one (and pretty much all the cards I get from every patient, if they have a kind note in them).
The result of the above stellar behavior? I would do virtually anything to try and help her. Back in the late 1990s, when she was very sick with AIDS, one of my ID fellows said it best: “There’s just something about her that makes you want to take care of her.” Exactly.
Based on patients like this, I was drawn to this commentary over in JAMA entitled “Do Nice Patients Receive Better Care?”
It’s a thoughtful piece, with no firm conclusion on the provocative question raised by the title. It’s also a good addition to the whole, “variations in care” theme that has already been extensively studied related to race, sex, patient income, education, and other factors. Here’s the big finish:
Clinicians are human and subject to the influence of bias. Patient behaviors will have a clear influence. It is important to recognize this phenomenon, put it in the proper context, and develop strategies for dealing with it to ensure that professional standards are met. On the other hand, nice patients and patients with nice families probably do receive a level of care that is perhaps at times well above the professional standard. Pretending that this phenomenon is not so is probably not helpful, and raises the next question—is it wrong?
Wrong or not, I can assuredly tell you this: While we may not be able to prove that nice patients get better care, they certainly make being a doctor a whole lot more rewarding. And fun.
And I thank them for that.
July 5th, 2011
Unofficial CROI 2012 Dates: March 4-10, in Seattle
No, there’s nothing up yet (as of July 5, 2011) on the official CROI web site.
But someone was kind enough to send me this link that lists upcoming meetings in Seattle. Here’s an extract:
Note that only those events with green shading are “Definite.” And since these dates haven’t yet been confirmed by the CROI-meisters, it’s probably best to hold off on the hotel and flight reservations for now.
But odds are we’ll be heading to the Pacific Northwest in March — perhaps they’re on a ten-year rotation!
[Edit: Above link is now broken, but it’s still there under the “Quick View” option if you Google search, “CROI 2012 Seattle”.]
July 3rd, 2011
Proof That We Are Not French
In case you were worrying about fading American national identity as we celebrate July 4, did you see this detail on a recent E. coli O104:H4 outbreak from France?
More recently, at least 15 people in Bordeaux, in southwestern France, appear to have been infected with the strain found in Germany. Most of them have been linked to a day care center in Bègles, a suburb, where the victims apparently ate gazpacho garnished with sprouts.
I think it’s safe to say that among day care centers in the United States, the number serving gazpacho garnished with sprouts is approximately the same as those serving Chateau Lafite as the drink to accompany chicken fingers with fries.
Or should that read, “avec frites.”
June 24th, 2011
Reflections on Levofloxacin as it Goes Generic
With the news that a generic form of levofloxacin has just been approved by the FDA, some thoughts about this remarkable antibiotic:
- When it was first approved in 1996, levofloxacin was the first oral antibiotic that really covered all common causes of community acquired pneumonia. Strep pneumo, H flu, mycoplasma, legionella, chlamydia — check, check, check, check, and check. Doctors realized this, of course, and prescribed it like mad.
- Grave prognostications about the threat of pneumococcal resistance to quinolones invariably followed — but this never materialized to a sufficient degree to change practice, at least not in the United States. It has remained unusual to find one of these levofloxacin-resistant pneumococci in clinical isolates, and surveillance reports show that such resistance is surprisingly rare.
- However, rates of gram negative (especially) and Staph aureus resistance to quinolones just keep going up and up. Amazing, these bugs are smarter than we are! What a concept! These UTI guidelines cite “collateral damage” of using quinolones for uncomplicated cystitis — a wise move.
- These cautionary notes notwithstanding, some medical services think it’s mandatory that every patient receive at least one dose of levofloxacin prior to hospital discharge. I made that up, and have mentioned it before, but you get my point — this is still one popular antibiotic. See #1, above.
- As numerous other quinolones fell by the wayside due to safety issues, levofloxacin has remained overall quite safe. Temafloxacin, trovafloxacin, grepafloxacin, sparfloxacin, gatifloxacin — gone, but not forgotten. Trivia question for detail-obsessed ID types (which means all of us): Why were each of these pulled off the market? And what were their trade names? (Hint: several of the trade names sounded like they were lifted right from a science fiction novel.*)
- But even though it’s relatively safe, levofloxacin does have some serious side effects. Probably the worst of these is the tendon rupture/tendinitis issue — deserving of the dreaded “black box” — but I’ve also seen anaphylaxis, delerium, QT prolongation, photosensitivity, and many many cases of C diff. Bottom line is that this is a drug: we humans did not evolve to have levofloxacin coursing through our system. (Tell that to cardiologists about statins.)
- In vitro, levofloxacin is neither the most active gram-negative or gram positive quinolone. Those would be ciprofloxacin and moxifloxacin, respectively. But it hasn’t really seemed to matter much, has it? See #1 (again), above.
- Presumably, the availability of generic levofloxacin will eventually eliminate the levofloxacin to moxifloxacin swap often mandated by payors. Since our hospital has levofloxacin as its preferred respiratory quinolone, it has set up the peculiar practice of using levofloxacin during the hospitalization, then changing to moxifloxacin on discharge. This can’t make medical sense — they are not identical after all — so I for one will be glad to see this exchange come to an end.
*An experienced starship pilot, Raxar expertly navigated his Tequin-400 aircraft through the high mountains of Trovan. This was no easy task, as the planet’s distinctive craggy peaks were nearly completely obscured by thick clouds of Omniflox — the highly-toxic gas released from volcanic eruptions. Off in the distance, he could see the bright lights of the capital city Zagam. “Home at last,” he said out loud. But Raxar spoke too soon, for little did he know that this final leg of his journey would take him to worlds unknown …
June 19th, 2011
Abacavir Agonistes
The studies on abacavir and its potential association with increased cardiovascular risk have been inconsistent ever since the news first broke at CROI 2008. But recently the data have been swirling around so fast and furious that it seems appropriate to take out this famous Greek epithet.
A summary of some recent notable studies:
- An FDA meta-analysis presented at CROI this year of randomized clinical trials — which avoids selection bias — finds absolutely no association with MI. Note that the analysis includes the studies in this ACTG report, and more.
- By contrast, a newly published observational VA study — including 11,000 patients, mostly men — demonstrates that among HIV therapies, only abacavir is associated with increased risk of cardiovascular events.
- Up in Montreal, here’s another study (n=7,000) with a positive association between abacavir and MI, this time along with several other antiretrovirals (including, oddly, efavirenz — marking the first time an NNRTI has ever been implicated in CV risk, and frankly raising more questions about study result validity than making me worried about efavirenz). Journal Watch coverage here.
- Meanwhile, the original VA abacavir study — the first one to introduce renal disease as an important possible cause of “channeling bias” back in Cape Town — has just been published, and this one includes even more patients (19,000) than the VA study cited above (I presume there’s some overlap). The result? It shows not only no significant association of abacavir with MI, but that abacavir is associated with a reduced risk of cerebrovascular disease (stroke).
- The editorialist commenting on the paper (Sam Bozzette) examines the numbers carefully and — somewhat surprisingly in light of the study findings — writes: “Moreover, the trends tend to support the controversial notion of a differential effect of abacavir [on MI risk]…”
- And what about that protective effect of abacavir on stroke seen in this study? A Danish study found just the opposite: on investigating possible risk factors for stroke among patients with HIV, the researchers showed that abacavir was associated with increased cerebrovascular events — and it was the only HIV treatment to earn this honor.
Got that?
Although it’s tempting to revert to the last refuge of the researcher, grant writer, and journalist — the generic “more research is needed” — I’m pretty sure we’ve reached a point of diminishing returns on these abacavir-cardiovascular disease studies, at least those of retrospective observational design.
And from a practical, day-to-day patient management perspective, in my view nothing really has changed despite these recent studies. It seems advisable to avoid using abacavir in patients with high cardiovascular risk if there are suitable alternatives (which there usually are), but that stable patients already on the drug should remain on it unless there’s a compelling reason to switch.
June 15th, 2011
Hockey Helicobacters
Today’s ID/HIV items come to you courtesy of a winter game being played during a summer month:
- So it appears that community-based care of HCV augmented by telemedicine is just as good as traditional clinic visits to specialists. My first thought on reading this important paper is that there are undoubtedly lots of ways to incorporate technology into patient care for the better, extending the reach of specialty services. But — and call me a cynic — if in a fee-for-service world, specialists get paid for services rendered during office visits, and not for setting up and managing telemedicine, which one do you think they’ll choose? To quote the editorialist: “It is also important to develop models for financing this innovative care model, with respect to both the specialists and the primary care providers involved.” Emphatically agree.
- Here’s a comprehensive list of sprout-related outbreaks, if you’re keeping score.
- In the flurry of recent drug approvals in the ID/HIV world — ceftaroline, nevirapine XR, rilpivirine, fidaxomicin [update: apparently not available until later this summer], boceprevir, telaprevir, generic zidovudine/lamivudine — I always wonder what the early anecdotal experience will be from experienced providers. Any first impressions? So far I’ve used ceftaroline and rilpivirine, not the others.
- Did you see this latest “scorecard” on states’ compatibility with the 2006 HIV Testing Guidelines? It includes three suggested key elements of HIV consent to facilitate testing: 1) changing from opt-in to opt-out 2) allowing general consent for care to include HIV testing, and 3) permitting written or oral consent. Good news: now only 4 states per this report have HIV testing laws outside these recommendations. In fact, there is only one state that is incompatible with all three components. And that state is … Massachusetts, thank you very much! (FYI, we’re working on changing this.)
- Can’t believe I missed commenting on this remarkable paper in Lancet ID, which clearly documents the clinical relevance of transmitted drug resistance, as well as the importance of baseline genotype testing in reducing the risk of treatment failure. It also hints that in the presence of any transmitted resistance, a boosted PI-based regimen might be the best choice, at least for virologic outcomes. Makes sense. Further commentary in Journal Watch AIDS Clinical Care here.
And as you watch tonight’s Stanley Cup final, and thoughts turn to winter, you might note that the 2012 CROI dates and location still have not been announced. Are the conference organizers awaiting the outcome of tonight’s game to decide where to go? Both Boston and Vancouver in February would be suitably cold options.
Paul E. Sax, MD
Contributing Editor
NEJM Journal Watch
Infectious Diseases
Biography | Disclosures | Summaries
Learn more about HIV and ID Observations.
NEJM Journal Watch — Recent Infectious Disease Articles- Simultaneous vs. Sequential Administration of COVID-19 and Influenza Vaccines
- Observations from ID and Beyond: Musings About a Bruising and an ID Link-o-Rama
- The Bad with the Good: Donor-to-Recipient Transmission of Bartonella quintana
- A Promising Novel Treatment for Progressive Multifocal Leukoencephalopathy
- Largest Recorded Blastomycosis Outbreak in the U.S. Occurred in Michigan