January 27th, 2019

For Our Stable HIV Patients, Why Are We Still Sending All These Lab Tests So Often?

Interesting query from a colleague recently:

I’m a community ID doc in the trenches (the measles trenches at present) with an HIV question. Why do we still check CBCs & chem panels every 3-6 months in our HIV patients? Particularly our well-controlled, virologically-suppressed patients? This strikes me as a tremendous waste. I haven’t been in practice that long, but I can count on one hand the number of times these routine labs have led to a change in ART (and even then, it was probably a patient on TDF, which I don’t use much any more). Is this an evidence-based practice? Or a vestige of an earlier era of more toxic drugs?
Thanks!
Andrew

Andrew raises an important question — do the guidelines for laboratory monitoring still make sense when our HIV treatments have become so safe and effective?

Below I’ve summarized the labs recommended by the DHHS Guidelines for our stable patients — the people who have been virally suppressed on ART for years. In italics, a bit of commentary.

  • CD4 cell count — if CD4 < 300, every 3-6 months; if CD4 300-500, every 12 months; if > 500, optional. Wow, that’s complicated. How about never? One could easily argue that CD4 monitoring is only needed in those very rare patients who have persistently low CD4 (e.g., < 200) despite long-term viral suppression. Remember, HIV treatment should not be changed based on CD4 results alone.
  • HIV RNA (viral load) — every 6 months. While the viral load is absolutely critical for monitoring adherence, does a strategy of twice-yearly measurement make sense for patients who have been on effective treatment for a gazillion years, are on a perfectly good regimen, and have never failed therapy?
  • Basic chemistries, LFTs, and CBC (when measuring CD4) — every 6 months. This was the group of tests alluded to by Andrew in the above email, and I agree that they rarely pick up something of HIV- or ART-related concern with current treatments. Exceptions would be for patients with known comorbidities — but this is a different indication for testing. About the only good thing you can say about this testing is that it’s cheap, at least compared to CD4s and viral loads.
  • Urinalysis — yearly if on TAF or TDF. Certainly this makes sense with TDF, and again also when there are concomitant risk factors for renal disease (diabetes, hypertension). But in an otherwise healthy young person not on a TDF-based regimen? The incidence of clinically important abnormalities with this yearly screen must be extraordinarily low.

Here, then, is a revised (and deliberately provocative) recommendation for monitoring the otherwise healthy people (who happen to have HIV), and who have long-term viral suppression — let’s say at least 5 years < 200, just to be safe. Let’s also assume they are also currently receiving a recommended regimen that does not include TDF:

  • HIV RNA (viral load) — yearly.
  • CD4 cell count — never.
  • All other tests (chemistries, renal function, LFTs, CBC, urinalysis) — at comparable age-appropriate or comorbidity-appropriate frequency to HIV-negative people.

When I’ve floated this idea by certain colleagues, they frequently cite the asymptomatic sexually transmitted infections they’ve picked up in their twice-yearly (or more frequent) monitoring.

I’d argue that this reflects an individual’s STI risk, which is not the same in all people with HIV. By all means, continue to screen for STIs when clinically indicated, and the same goes for underlying medical problems that increasingly arise during aging.

So to test this revised strategy, let’s imagine a clinical study:

Eligible: Stable on guidelines-approved ART; no history of virologic failure or treatment interruption; HIV < 200 on all measurements during the past 5 years.
Intervention: Randomized to 1) guidelines-recommended monitoring, or 2) HIV RNA once-yearly, other testing as indicated by demographics, clinical status, comorbid conditions, STI risk.
Primary endpoint: Virologic suppression at the end of the study.
Secondary endpoints: Occurrence/diagnosis of HIV or non-HIV-related comorbidities; cost (to healthcare system); cost (to patient).

And, since we’re talking about a way to reduce office visits and healthcare utilization, how about this recent “appointment” in Alaska?

12 Responses to “For Our Stable HIV Patients, Why Are We Still Sending All These Lab Tests So Often?”

  1. ccg says:

    Here’s another angle. PEPFAR now supports only very limited CD4 testing even at initial diagnosis which means advanced disease is diagnosed on clinical grounds. The important downstream effect is that it is hard to identify the patients who need prophylaxis for pcp and diagnostics for cryptococcal disease. This is an active discussion in the international treating community. Supporting CD4 testing is hard. VL is better at identifying treatment success and is being actively scaled up-for so many reasons. It is a conundrum in the setting of limited resources and a fixed budget.

  2. MN MD says:

    Serena Koenig et al. looked at similar question in the Sept 1, 2010 issue of Clinical Infectious Diseases. Perhaps she has longer term follow up or is game to look at same in a Harvard cohort.

  3. Loretta S says:

    This is just so darn sensible, Paul. Not to mention economical. And also not to mention patients who will have to get a blood draw less frequently and will be happy about that.

  4. Jeff Kirchner says:

    I generally agree with what Paul has to say but with a few caveats and coming from my recall in the “old” days when we were doing these tests monthly.
    – for aging patients a different set of rules due to growing # of comorbidities and medications that still need monitoring of hepatic and renal function – so would continue to believe that q 6 month CMPs are reasonable. I am willing to give up the CBCs . Regarding CD4 counts – I see the logic of “never” after they are normalized but again with aging patients who are often hospitalized (and maybe never attained full immune recovery) it if often clinically helpful to know if the most recent CD4 count was 150 (like a few of my older stable patients) or 550. I sometimes see patients of a colleague without a CD4 count for > 2 years and then checking when clinically ill – as we learned many years ago, can be misleading.
    Last point – many of my older patients do NOT want to even come in to the office any longer “I am doing fine” – so we have negotiated with yearly visits and q6 month VLs and CMPs (and other clinically appropriate labs – A1C, lipids, ).
    One more clinical anecdote, 6 month visit last week “stable” female patient (16 years since DX) – she looked well in the office “Yes, I am taking my Genvoya every day” — viral load at this visit 86,000 copies – I think probably not taking. Even our long term patients need support / adherence counseling.

  5. Teddy says:

    Very interesting discussion. I think CD4 is still key in a context where the diagnosis of HIV is made late. Have a number of patients dying because of IRIS and it can help in close monitoring. Remember the prevalence of HIV is highest in areas with fragile health systems- sSA.

  6. Wendy Armstrong says:

    Here is another twist : ADAP and HICP in Georgia require CD4 counts every 6 mos, even for our patients with very high CD4s. I am also required to order annual RPRs and GC/CTs on Ryan White pts even for those who are not sexually active. This is not “Choosing Wisely” and wastes a lot of money for unnecessary testing. We need to advocate to change these requirements.

  7. Harley Cox says:

    It is not so economical when, with reduced or no testing of viral load, the partners of PLWH would need to take Prep rather than rely on TasP.

    This model does not only create savings but very real costs.

    I’m a PLWH who had a massive treatments failure, with a viral load in the tens of thousands. This was picked up by a regular blood test.

    I’m not at all certain that presenting at an ER with an AIDS defining condition is a positive outcome either for the patient, their partners or the ER staff.

  8. Gabriel Cuba says:

    In Brazil, our guideline made some of these suggestions mandatory: 1) we are not performing CD4 count if persistently above 350, viral load indetectable and any severe comorbidity (cancer etc); 2) basic chemistries etc every 6 months; in fact, stable patients returns to medical visit twice an year.

  9. Leslie Linares-Hengen says:

    Hi Paul,
    Thanks for your insight as always. I have tried to get patients used to the idea that they don’t need forever CD4 monitoring. Most eventually are OK with idea. However, those on drug assistance programs have forms that need to be filled out every 6 months which always ask for CD4 count in addition to viral load. This has kept me checking more frequent CD4 counts that I would have liked in stable patients.
    Regards,
    LALH

  10. Serena Koenig says:

    I read this blog and all of the comments with great interest, because we have focused a great deal on this topic in Haiti. We have a very limited budget, and need to use our resources in the most effective manner possible.

    We did the study that was cited in comment #2 above (thanks for reading/remembering our study). The reason we did the study is exactly the same reason that Andrew described in Paul’s blog above. We felt like almost all of the routine chemistry and CBC tests that we were ordering were normal, and it seemed like a poor use of resources. For this reason, we conducted a retrospective study to look at the clinical impact and cost of each test. We found that most of the test panels were of low yield, so we stopped the policy of routine testing, and we shifted the money to better uses.

    To me the question is what kind of research is necessary to answer the clinical question that Paul describes so eloquently above, and what is necessary to change the requirements for twice-yearly CD4 counts that Wendy describes. I know that Wendy is very focused on the great needs of her patients in Atlanta – such as the costs of getting to clinic. Is there any way that those financial resources could be re-directed from frequent CD4 counts to transportation subsidies?

    I would be very interested to see a study like Paul describes above done in the USA – or even as a start a retrospective analysis of lab tests, with a determination of the proportion of tests that were abnormal, and impacted clinical care.

    I really appreciate everyone’s thoughts on this very important topic.

  11. Darcy Wooten says:

    I completely agree! Less is more! But…as was mentioned above, many funding organizations requirement more frequent testing. This can likely change but until it does, I imagine that providers will still feel the need to “over test” in order to satisfy these regulatory bodies. Like most things, it always comes down to money. Having to practice in accordance with funding and regulatory bodies is just one of the many things that continue to take the joy out of medicine. Sigh. 🙂

    • Mark Boyd says:

      This series of comments reflects thoughts I have also had over the past couple of years. I have stopped mindlessly checking CBC, UEC and LFT in patients without clinical indication (e.g. comorbidities). I have also stopped checking T cell subsets in people with persistent undetectable VL and CD4+ >350. As per Paul, I generally only test STIs in those at risk or symptomatic.
      Another way to think of this issue is to ask what evidence supports monitoring at all? It is of course time-honoured, but is there any solid evidence that it makes any difference to hard patient outcomes?

Leave a Reply

Note: This is a moderated forum. By clicking on the "Submit Comment" button below, you agree to abide by the NEJM Journal Watch Terms of Use.

Our physician bloggers cannot respond to requests for personal medical advice, and recommend patients discuss health issues with their individual physicians.

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.