August 23rd, 2015

Post-Exposure Prophylaxis for HCV Can’t Be Cost-Effective — But We Might End Up Recommending It Anyway

An email query from a colleague:

Hi Paul,
Just got a call from one of our surgeons who got a needlestick from a suture needle, small amount of blood. Patient is HCV +. Any post-exposure prophylaxis recommended?
Thanks,
Dan

The quick answer is no, it’s not recommended. From the guidelines:

MANAGEMENT_OF_ACUTE_HCV_INFECTION___Recommendations_for_Testing__Managing__and_Treating_Hepatitis_C jpg

But it’s a natural question to ask for several reasons — on first blush, PEP for HCV seems like a no-brainer, because:

  • It works for HIV.
  • There’s no vaccine.
  • Rate of transmission is approximately 10-fold higher than with HIV.
  • The HCV drugs are practically side-effect free.
  • It feels better doing something rather than nothing.

But despite the above issues that favor PEP today for HCV occupational exposures, there is absolutely zero chance this makes sense from a cost-effectiveness perspective. It’s just not possible.

A little refresher on basic principles of cost-effectiveness analysis:  something is cost-effective if you get good value for your healthcare dollar. You spend more money, and you prolong life. How much you spend compared to standard-of-care, and how more much life you get in return gives you the numerator and denominator in that all-important cost-effectiveness ratio. In the USA, if we spend <$100,000 for an extra year of life, that’s considered cost-effective; <$50,000, very cost effective.

So let’s consider the HCV needlestick scenario, with 100 exposures; all assumptions will be biased in favor of PEP, just to make the point:

  • The risk of acquiring HCV from a needlestick is 2-8%, depending on the nature of the exposure; some will clear it spontaneously. So if we do nothing, around 5 of these 100 exposures will end up with chronic HCV.
  • Let’s assume that two weeks of PEP with sofosbuvir/ledipasvir will be 100% effective in preventing HCV acquisition. There are no data, but let’s give it the benefit of the doubt. And note that the HIV PEP protocol is 4 weeks, but at least one pilot study is using 2 weeks for HCV. (Why? Not sure.)
  • Two weeks of PEP will cost $15,000 per patient — actually, a bit more, but let’s round it to $15K or a total of 1.5 million dollars for our 100 exposures.
  • If we didn’t give PEP after the exposures, we’d need to treat the 5 people who contracted HCV with 12 weeks of sofosbuvir/ledipasvir, which would cost $450,000. All would be cured.
  • The PEP strategy therefore costs $1,050,000 more than the no PEP one (1.5 million minus 450,000), and prevents 5 cases of HCV. That’s $210,000/case prevented — so a lot.
  • What about survival, the denominator part of the cost-effectiveness ratio? The projected survival of those in whom HCV was prevented by PEP vs later cured with HCV treatment should be roughly the same (no one is getting cirrhosis or hepatocellular carcinoma with early treatment) — but maybe there’s some “disutility” to getting HCV, even if it’s rapidly cured. So let’s give the PEP strategy a bit of a quality-adjusted survival advantage — how about 10 weeks, since that’s the extra time on treatment for an actual case? Of course we can only apply this to the 5 cases of HCV prevented in our 100 person cohort, so it’s an overall advantage of 0.5 weeks (10 weeks times 0.05), or roughly 0.01 years.
  • Finally, we arrive at our cost effectiveness ratio, which is $1,050,000/0.01 years, or $105 million/quality-adjusted life year.

More than 100 million dollars for a quality-adjusted life year? Yep. I told you it couldn’t possibly be cost-effective.

Now again, I biased everything in favor of PEP. What if it doesn’t always work? What if failures of PEP lead to resistance? What if the PEP regimen needs to be 4 weeks long? What if the quality-of-life adjustment is an overestimate? What if some of the patients who get HCV have a very low HCV RNA, and can be cured with 8 weeks, not 12?

Well, then you’d be spending even more than $105 million dollars per year of life saved. But whether it’s $105 million, or even half my back-of-the-envelope estimates (which could be wrong — I was an English major, after all), the point remains: This can’t be a cost-effective intervention.

But as I was sharing these thoughts with my brilliant friend and colleague Rochelle Walensky, who does this stuff for a living, she reminded me of this important fact:

I might encourage you to acknowledge that many of these kinds of decisions (PEP, safety of the blood supply) are not made on cost-effectiveness grounds … Most of the cost-effectiveness data for the screening for the blood supply demonstrates it’s not at all cost effective. You may want to acknowledge that point (we pay more for lots of things that cost-effectiveness suggests we shouldn’t, including keeping health workers safe given their dedication to care, keeping blood supply safe etc).
Acknowledged — I told you she was brilliant. Which is why, despite the fuzzy math above, it wouldn’t surprise me a bit if in the not-too-distant future, we are actually recommending PEP for HCV.Nice 6 minute lecture here. Enjoy!

[youtube http://www.youtube.com/watch?v=RSBaEz10sQk&w=560&h=315]

10 Responses to “Post-Exposure Prophylaxis for HCV Can’t Be Cost-Effective — But We Might End Up Recommending It Anyway”

  1. Thanks dear Paul for this interesting and totally justifiable discussion. However, I believe it’s not a matter of white and black!
    For situations like transfusion of a pack of HCV positive blood or blood products, although rare, but possible, specially in emergency situations, we should think outside the box and accept it. In this situations it is even cost- effective too!

  2. Mai Tuyet Pho says:

    There’s Uptodate for knowledge, and HIV and ID Observations (and Rochelle Walensky) for wisdom. Thanks Paul!

  3. Nick says:

    Good article Paul. The cost-effectiveness analysis paints a good picture of how to think about cost in a simple and easy-to-follow manner

    Regarding Rochelle’s comment, I think there is a key consideration for when to use cost-effectiveness data that has been omitted. For the examples she provided, would the outcomes for the patients be the similar depending on which strategy is chosen? Most likely a patient receiving PEP for HIV would have significantly better outcomes than a patient not receiving PEP who would require chronic treatment if they did in fact contract HIV. However, in the case of HCV, outcomes between PEP and no PEP may actually be fairly similar with the very high cure rates seen in patients treated with the newer agents. Thus, this is probably a very good example of when cost-effectiveness research could potentially help to guide therapy decisions.

    • Paul Sax says:

      Nick, this is an excellent point — which is why giving a survival advantage to those who get PEP for HCV was kind of forced in my example. There probably is no survival advantage, so the cost/QALY could be infinite!

      Paul

  4. Erika says:

    Would the stuck people actually receive treatment? Or would the 5 infected people need to wait until they had symptoms or severe symptoms to receive treatment?

    If they did, it would increase the cost-effectiveness … but would feel bad. Supports your final point maybe.

  5. Tim says:

    Thanks for the discussion! Along the lines of cost effectiveness and HCV, I’m having a tremendous problem getting insurance companies to approve the new HCV therapies. Most want evidence that the patient has advanced disease. I thought the point was to treat people before they developed complications?

    How are you handling the pushback on these drugs from insurers?

  6. Erin says:

    Thank you for a wonderful article. My colleagues and I are wondering if you have the reference for the 2 week pilot trial you mentioned? Thank you!

    • Paul Sax says:

      Erin,
      The PEP trial is being led by Ray Chung at MGH — would check with him. It’s not on clinicaltrials.gov yet, but will be.
      Paul

  7. Rebecca says:

    Hi Paul,

    I work at the Clinician Consultation Center at UCSF where we provide phone consultation for healthcare providers managing exposures to HIV, HCV, etc.

    Your blog post already made it’s way around the office, bravo. We get this question at least a few times a day so your analysis gives us fun numbers to talk about when we tell callers, “nope, no PEP for HCV, sadly it would cost too much money BUT the new meds would cure you if you get an acute infection.”

    When you said, “The risk of acquiring HCV from a needlestick is 2-8%”, we were wondering where the 8% statistic is from. Shared needle exposures?

    • Paul Sax says:

      Hi Rebecca,

      Thanks for your note. If one scours the literature for all the various case series on HCV transmission after needlestick, there are some that have rates this high; here’s a small one with a 10% rate:

      http://www.ncbi.nlm.nih.gov/pubmed/1427651

      Certainly lower with a solid bore needle. And the guidelines cite 7% as the upper limit, sorry if I disrupted that assumption — you can stick with that. (Pun intended, apologies!)

      Paul

HIV Information: Author Paul Sax, M.D.

Paul E. Sax, MD

Contributing Editor

NEJM Journal Watch
Infectious Diseases

Biography | Disclosures | Summaries

Learn more about HIV and ID Observations.