AMCHP 2005 ANNUAL CONFERENCE
DELIVERING RESULTS, IMPROVING PREGNANCY & BIRTH
February 19-23, 2005
MARGARET LAMPE: Thank you, Sharon, for inviting me to give this talk. I really -- I work in the Division of HIV, it's prevention, and it's really a lot of fun for me to connect with my MCH colleagues, I don't get to do that as much as I would like.
First, I'm going to just sort of give sort of a national overview basically of our guidelines focusing mostly on counseling and testing because that's where we find we still have room to improve.
This first slide is a representation of perinatally acquired AIDS cases by year of diagnoses in the United States. Now mind you that AIDS is not HIV infection, but rather children who are HIV infected who progress to AIDS and in the era of highly active antiviral therapy that can take a much longer time. And the reason we report AIDS cases is that our HIV reporting is incomplete throughout the country.
But this slide demonstrates the tremendous decline in perinatal AIDS cases from a peek around 1992 and our newest data is from 2002. We would like to get to zero on this graph. We would like to get to zero actual HIV infections.
this will probably be a review for many of you, but I wanted to just reiterate what was reported in the Institute of Medicine report regarding perinatal HIV prevention from 1998. Basically the chain of events that lead to an HIV infected child or HIV infected children are the proportion of women who are HIV infected who become pregnant who do not seek prenatal care who are not offered HIV testing, those who refuse testing who are not offered antiretroviral prophylaxis, who refuse antiretroviral prophylaxis, and I would actually insert here for women who don't have a viral load of a thousand or less are not offered or declined a Cesarean section. That's a research development since 1998 that has shown to be effective. And also women who --
UNIDENTIFIED PERSON: You can't be heard in the back of the room. That's why I moved up front.
MARGARET LAMPE: Oh, okay. Is that better?
UNIDENTIFIED PERSON: Yes.
MARGARET LAMPE: Thanks. Those who do not complete the antiretroviral prophylaxis, and then lastly those whose children are -- whose child is infected (inaudible) there's still some treatment failures.
The Francois Fabiet Beck Youth Center sort of adopted this cascade to describe some interventions that are particularly applicable to state MCH and those steps in reducing perinatal HIV transmission is good primary HIV prevention in women.
Secondly, prevention of unwanted pregnancy in HIV infected women. Preconception counseling and care. Accessible affordable welcoming prenatal care. Universal prenatal HIV testing. Reoffering to those who declined or at risk. Providing antiretroviral prophylaxis, and again I added and C-section to all eligible women, support for the parents to antiretroviral prophylaxis, it can be complicated and difficult at times. And then rapid HIV testing for women with unknown HIV status, and I'm going to talk a little bit more about that. Antiretroviral prophylaxis of exposed newborns and of course comprehensive services for mothers and infants.
I want to just mention we, in April of last year, CDC announced a new initiative called Advancing HIV Prevention, that's what the HAP stands for. And there were four primary strategies, and among them was to further decrease perinatal transmission. Myself, I was very excited to see that included, because frequently in an HIV organization, it's very hard to get perinatal on the agenda so we were very happy to see this. But some of the priorities were to work with partners to promote routine voluntary prenatal testing with the option to decline, that means opt out testing. To develop guidance for using rapid testing labor and delivery or postpartum and promote its implementation. To monitor the integration of routine prenatal testing and rapid testing at labor and delivery and into medical practice. And then also conduct a case control study to assess why perinatal infections are still occurring, in particular the treatment failure I mentioned.
This is somewhat untimely, I know folks are relooking at newborn screening, but I borrowed this slide from the New York City Department of Health. I modified it a bit with the HIV data, but these are conditions that are routinely screened for in New York as part of newborn screening. And as you all know, conditions that generally make this list are those that are serious and preventable or have a good treatment available. And HIV certainly meets that criteria. And as you notice, the prevalence of HIV exposure in the US, it's actually more so in New York, is much greater than any of the other conditions routinely screened for in newborn screening.
Just to give us some structure from which to operate, I want to describe some policies. About 35 out of the 50 states have specific laws or regulations regarding prenatal HIV testing, and they generally fit into one of three categories.
The voluntary approach is what CDC supports they either fall into an opt in or opt out category, and quite honestly these are sort of simplistic terms for the State laws, but most of them fit into one of these two categories. Opt in generally requires pretest counseling and written consent specifically for an HIV test. So if a women comes to prenatal care or presents to labor and delivery, she would get counseling around HIV transmission and her risk and would be offered HIV testing with specific informed consent for a test.
Opt out, on the other hand, treats HIV as more routine like other prenatal labs, and women are informed that she'll be tested for HIV and that she has the option to decline if she doesn't want to.
Lastly the mandatory approach, there are new states that have mandatory prenatal testing or mandatory testing of pregnant women, but two states, Connecticut and New York, have a mandatory newborn screening law, which means infants are tested with or without mother's consent when the mother's HIV status is unknown at delivery.
In 2001, CDC looked at our recommendations and put together -- looked at the existing data that we had to refine our prenatal testing recommendations and there were three primary data source that we used. Firstly, some chart reviews in eight states from a random sample of prenatal and labor and delivery charts.
We also use PRAMS, the Pregnancy Risk Assessment Monitoring System, there were nine states that participated in PRAMS and asked women postpartum if they were tested for HIV during pregnancy.
And then lastly we looked to our neighbors in the north, Canada, they have all of their HIV testing is done in provincial labs, and they actually have very excellent data around prenatal testing. So we looked at those three data sources.
And when looking at the medical record review to compare the different policies that were in the State, there was only one state that participated in the record review that had an opt out policy, that was Tennessee. As I mentioned, New York and Connecticut, they actually switched to mandatory testing around this time. New York's test law was modified to increase to require a turn around of results earlier than it had been previously, so that's why you see the difference, the two mandatory laws for New York.
And what we found, these are prenatal HIV testing rates to the far right and what we found was that mandatory and opt out policies, mandatory newborn testing and opt out policies for women achieved higher -- much higher prenatal testing rates as one would expect. In fact, in this example, opt out testing had the highest rate of prenatal testing.
Also from that study we wanted to look at HIV versus other routine perinatal infections. And Hepatitis B, syphilis and rubella all had documentation of test results well above 95 percent, whereas HIV only had around 52 percent of the records had test results documented.
When looking at the PRAMS data, again, there was only one state that had opt outlaw, that being Arkansas down here towards the bottom of the middle, and then New York again was in this sample in PRAMS. And again, opt out and mandatory testing, mandatory in New York had the highest rate, but it surpassed opt in testing in most instances.
And then most convincingly, though Canada is another country, it's somewhat similar to the US and they actually had the best data when comparing two provinces, Alberta and Newfoundland, to the three provinces that had opt in, hose difference in testing rates were quite stark. In fact, roughly around 95, 98 percent for opt out, whereas opt in was significantly lower, 54 to 83 percent.
So some conclusions that we draw from this is that we really needed better data in the US to make some judgments. But we felt like we had enough evidence to make an opt out recommendation for prenatal testing and our developing a randomized record review methodology that is currently under way in South Carolina and will be continued in eight other states currently. And then once that has been fully piloted and done in several states, will make that methodology available to all states.
Now, I sort of jumped ahead, but I'm going to go back to where we are with our testing recommendations and historically it's been a balance of risk benefits as many screening recommendations are.
Our first set of guidelines for perinatal testing in the United States were in 1985. This was a period, if you remember, when there was no treatment available. HIV was really identified in 1983. It had been determined that it was perinatally transmitted and it was a stigma that people living with HIV or diagnosed with HIV faced. So at that time, the general opinion was that the risks slightly outweighed the benefits of learning your HIV status if you were pregnant.
We published a second edition in 1995 and this is after the clinical trial called the AIDS Clinical Trial Group 076 which demonstrated that AZT given to the mother prenatally intravenously, antepartum and to the newborn orally for six weeks reduced perinatal transmission by about two thirds. And just, most of you are probably familiar with this. Without any treatment, roughly 25 percent of newborns are born infected. With AZT, the three arms of AZT, about eight percent. And that was the State of the art in 1995.
So at that time, we recommended universal counseling and voluntary testing such as an opt in approach. Many states have modeled their laws after the '95 recommendations. And after that time, we felt we did see a marked decline in perinatal cases attributed to screening and the use of AZT.
And then in 2001, as I say, we looked for the data for the MWR, but prior to that we had published updated recommendations. And at this time, the maternal -- there were maternal treatment advances, heart therapy or highly antiretroviral therapy were really available around 1997, 1998. It was demonstrated to really prolong the lives of many individuals. So we recognized that with the maternal treatment advances, both the mother and the baby would greatly benefit from her learning her HIV status. And at that time, we recommended that HIV screening be a routine part of prenatal care. We also recommended routine testing in the third trimester for women who are at risk and who live in high prevalence areas.
And then lastly, rapid HIV testing for women in labor with unknown HIV status.
And then in April of 2003, we further enhanced that recommendation to support more strongly the opt out approach, and this was a colleague letter that
Dr. Gerberdean, CDC director, sent to all the HIV and MCH directors and obstetricians in the country. And basically that dear colleague letter said no child should be born in the US whose mother -- the child's HIV status or its mothers status is unknown, promoting routine opt out screening prenatally, rapid opt out testing at labor and delivery. And then with the newborn testing, we just sort of left that to a state law. And you can find -- this web site at the bottom is where you can find this dear colleague letter.
Since then, we work with our colleagues at ACOG, and they published and updated and expanded recommendation for perinatal HIV testing through their committee opinion which is how they generally publish their guidelines and recommendations. And the simile of that is it's very similar to our CDC guidelines which are opt out prenatal HIV testing, repeat testing in third trimester for women in areas with high HIV prevalence who are known to be at high risk for infection and also who declined earlier HIV testing, that's an addition that we don't have. They also recommend rapid HIV testing for women in labor for women with undocumented HIV status and they go forward and recommend, because they are the clinicians, to initiate antiretroviral prophylaxis with consent for women without the results without waiting for confirmatory results. And this is a big change in clinical practice.
So I want to just focus a bit on rapid testing because this is sort of the newest technology and the newest change. We've been promoting and trying to do prenatal testing for a long time as you see. So why would we want to offer rapid testing for women in labor? Well, in 2000 we estimated that 6- to 7,000 HIV infected women gave birth in the US, and that resulted in about 280 to 370 infants. As I said HIV reporting among exposed infants and infected infants is somewhat incomplete. This is our best estimate. As you see it's in 2000. We are working on updating this to get more accurate, current data.
But of those 280 to 370 infected infants, about 40 percent of them were born to only with unknown HIV status prior to delivery. So 40 percent weren't treatment failures, they were unknown and unavailability to access any interventions. And with a 25 percent transmission rate, you could estimate that three times this number were not infected due to chance alone.
More to the rationale. Labor and delivery is really an opportunity. It's a 48-hour window where women -- most women in the United States who give birth do give birth in a hospital and frequently this is their one and only access with the healthcare system. Also, there are four FDA approved rapid tests that are now available. This is a rapid, no pun intended, change. The first test was the SUDS assay which some of you may be familiar with. Its performance wasn't the best, if you will, and people grew a little leery of rapid testing, but these four tests had very good sensitivities and specificities.
The rational for rapid testing in labor and delivery, as I mentioned, the estimate transmission rates without any antiretroviral therapy is approximately 25 percent. With the probably with antiretrovirals during labor, you can reduce the transmission rate down to about 9 to 13 percent, so about a 50 percent reduction. Although if you identify a women early in pregnancy and can do the antepartum, intrapartum and postpartum regimens, you can achieve 2 percent or less. And for those women who achieve a viral load of less than a thousand, a C-section isn't necessary, but if you can't get her viral load down that low, a C-section provides added benefit to get you this low to less than 2 percent.
Some evidence that we have, Sharon mentioned the MIRIAD study, this is the mother infant rapid intervention at delivery study, the objective of this project were to determine the feasibility and performance of rapid test for women with rated undocumented HIV status, to provide timely antiretroviral drug prophylaxis to reduce perinatal transmission, and to facilitate follow-up care for HIV infected women and their infants.
We conducted the MIRIAD study in 16 hospitals in these six cities, and this -- I'll quickly go through this. Roughly -- this is data through June of 2003. We are updating this analysis, we just finished enrollment into MIRIAD last month, actually.
But from November 2001 through June 2003, roughly 90,000 encounters were evaluated at 16 hospital L and D units. The reason we say encounters is because sometimes women would come more than once. If anybody has worked in a L and D you know we frequently have frequent fliers we like we like to call them. Usually they are the worried -- worried well. There were 7,000 -- roughly 7,000 women who were eligible to participate in the study, and to be eligible they had to have no HIV results in their record and were at least 24 weeks gestation. We picked that because of the viability of the infant. And then 78 percent of those women were approached and offered MIRIAD, and 84 percent of them consented for participation in our testing. And the reason for the what I consider to be somewhat of a lower approach rate is that in some hospitals they had funded staff only 9:00 to 5:00 to approach women to participate in the study. And this acceptance rate I think is also slightly lower than one might expect in clinical practice because they are consenting to participate in research to have many specimens drawn and other types of things. In MIRIAD we use the oraQuick rapid HIV test using an investigational device exemption, and what we found were -- excuse me. There were four false positives, and we tested women using the oraQuick. We also used what the current gold standard was, the Eliza or the EIA and there were 4 false positive results for oraQuick. However, there were 11 false positive using the gold standard, if you will, EIA. There were no false negatives, sensitivity or oraQuick was 100 percent. The specificity was 99.9 percent, and these are the 95 percent confidence intervals around the point estimates.
And then the positive predictive value for oraQuick was 90 percent whereas EIA was 76 percent. The reason I present this, this is really what folks are concerned about in screening a low prevalence population that they'll have false positive results and initiate prophylaxis unnecessarily. But this was a relatively rare event and this was -- for screening test standards, this is actually quite good.
The turn around time, these block plots are the median, and of course the Cortel ranges for the turn around time for rapid testing, the turn around time was 70 minutes, whereas the EIA was 28 hours. So if you want to get antiretroviral prophylaxis in before a women delivers, these timely results are really important.
There was quite a range as you see with the rapid test and the EIA, but we really are trying to encourage hospitals to achieve 60 minutes or less as a goal. The test itself, the oraQuick test requires 20 minutes to read the result.
Some difference in turn around times was where the test was performed. Point of care testing new the turn around time, and by point of care I mean that the nursing staff or physicians did the testing themselves in a labor and delivery unit in 45 minutes, and lab based testing was three and a half hours. And so our general recommendation is that consider point of care testing, particularly if they have an off-site lab or they don't have a good (inaudible). The oraQuick assay and the Unigold assay are creawaved and rather simple to perform. And we found that point of care testing delivered as accurate results as laboratory based testing.
So what were the lessons from MIRIAD? In laboring women with undocumented HIV status rapid testing using oraQuick delivered accurate timely results. I didn't present this data, but acceptance of testing in labor was high but varied by time and day of the week. Women were less likely to accept testing on the weekends. I don't know if it's the characteristics of those who come to L and D on the weekend. Testing performed at the point of care delivered more timely results and MIRIAD allowed previously unidentified HIV infected women immediate access to intrapartum and neonatal antiretroviral prophylaxis.
In 2002 the Office of the Inspector General did a report on barriers to HIV testing among obstetricians, and they issued a series of recommendation which was a list of tax for CDC eventually, one of the things that we developed a model protocol and facilitate a clinician's states rapid testing as a standard of care.
So what we did was we convened a group of folks with expertise in these areas that impedes public health at the rapid testing technology and so forth lab science to develop a model protocol. This is available on our web site. The purpose of it is to provide guidance to clinicians, laboratory and administrator, public health professionals and policy makers and it provides a general structure protocol that can be adapted locally. And within that document we recommend that hospitals adopt a policy of routine rapid HIV testing using an opt out approach for women who have undocumented HIV results when presenting to labor and deliver.
And in conclusion, until all pregnant women with HIV accessed screening prenatally, the promise of the clinical trial group 076 which I mentioned and the other clinical trials can't be realized. And rapid testing in labor and delivery really serves as safety net or that last opportunity to reduce the impact of missed prevention opportunities.
This is a list of resource, and I realize you can't jot these down if you're interested. If you would -- I'm going to pass a sheet of paper around if anyone would like copies of my slides, I'd be happy to E-mail them to you.
That's it.
UNKNOWN SPEAKER: (Inaudible.)
MARGARET LAMPE: Sure, that's fine. That would make it easier for folks, I guess.
Should I take questions or move on to Lois? Or what would you? I don't know if I'm over time. I'm assuming there are questions. Do you folks have questions or comments? Yes, sir
UNIDENTIFIED PERSON: One is a comment the other is a question. One is that California and Tennessee they have the opt out.
Two is the rapid testing. Is CDC recommendations for rapid testing on an outpatient basis in those areas that have high risk populations, high risk in terms of certain areas. For example in California, the border counties tend to have an increase, we don't know why exactly, but we have some speculation about that as to the HIV positive rates that may be going up.
MARGARET LAMPE: In the population in general or specifically pregnant women?
UNIDENTIFIED PERSON: Specifically in pregnant women.
MARGARET LAMPE: Interesting. I mentioned the advancing HIV prevention initiative and that there were four strategies, perinatal being one of them. The availability of rapid testing really is what has driven this new initiative, and we do have guidelines and recommendations for more routine HIV screening in primary care settings, and, you know, the talk around the office is that's the most widely ignored CDC recommendation--most widely ignored recommendation that CDC ever published. And we are trying to improve, particularly in care settings and community based organizations HIV testing. And we have demonstrative projects in California. I don't know if there's any on the border necessarily, certainly in San Francisco, but you're absolutely right. Rapid testing really provides a tremendous opportunity, not only in labor and delivery, but particularly out in the community where people don't access healthcare. When we look at the numbers of infected persons, I think about two thirds of them had accessed emergency room settings and had never been offered HIV testing, and so they live their lives and, you know, expose others. And so we feel that also instituting rapid testing in ER settings can help folks learn their risk status. Thanks, good question. Sharon?
UNIDENTIFIED PERSON: Margaret, are you following the roll out of the statewide implementation of rapid testing labor and delivery in Illinois right now? I know they are -- they interpreted their perinatal HIV laws and regulation that that would include having rapid testing in all their labor and delivery hospitals in
the state and how is that going?
MARGARET LAMPE: There are a couple of states that have taken a very proactive approach to really implementing this statewide. Illinois probably has the most hands on in labor intensive approach. Their state Health Department actually using very few CDC funds in terms of CDC funding terms, from their Health Department general fund, they are training every single hospital in Illinois that sees a pregnant women to implement rapid testing you using the oraQuick assay. I think it's going very well. I think they said in the last time I talked to them about a third to a little more than a third of the hospitals are already up and running. They've been at this about eight months actively. They had about a year of planning to get this up and going. You know, they basically wrote the law and then they went out to implement it. Their law really requires that women be offered testing in labor and delivery, rapid testing. And so they are out there to help the hospitals comply with their new regulations.
It's sort of like, you know, there's no -- one reason we developed the model protocol is we recognize all hospitals are their own cultures and communities and have their difference. So we realized there's not just one way to do this. Some labs, you know, some hospitals have a 24 hour lab that can turn things around very quickly where others don't. Some have a high turn over of L and D staff, so doing point of care testing would be really challenging and not possible whereas others have a higher nursing staff and could handle that additional responsibility.
UNIDENTIFIED SPEAKER: And also workers that are working on this project with Margaret, they are doing regional training for hospitals that are interested in implementing rapid testing in labor and delivery. And I was wondering if their thoughts are going to come out with recommendations of, you know, point of care versus lab testing or what are the elements that need to be in place or any kind of like lessons learned so if any states are thinking of implementing or in the process, whatever stage they are, to kind of have that document that would be so key to help with that.
MARGARET LAMPE: I feel the model protocol is -- the trainings are really based on the model protocol and if you're looking for a document, that's probably the best thing to look to right now. If fact, we are updating that right now and incorporating some of these lessons from the workshops. It's really a Johnny Appleseed approach. We don't have the funding to go out there and touch every hospital in the country, but we have invited -- we have eight workshops, and the last one was in Raleigh, South Carolina, covering the southeast. The next one is in Providence, Rhode Island, for the northeast. Basically what we do is invite high prevalence, high volume delivering hospitals that have a relatively high no prenatal care delivery rate that aren't doing rapid testing in L and D already. Those are the four primary criteria in which we select hospitals to invite.
We found it thus far a very effective strategy. We bring them together. It's sort of similar to an AMCHP action learning lab sort of model, where they have time to work individually, the lab director, OB director, hospital administration and L and D nursing supervisors really sit together at a table and build a strategic plan and an action plan about how they are going to make this happen. They hear some information from us and we really help facilitate heir internal planning. It's been a valuable tool. It's somewhat cost intensive, but we also try to pick institutions that others in the community look to for their -- you know, there's some old standard hospitals in the community and we are hoping this Johnny Appleseed approach will sort of spread throughout the communities where the key hospital are adopting his practice.
Yes, ma'am.
UNIDENTIFIED SPEAKER: I am from Bahamas and we have rapid testing in our hospital. So I was wondering at what stage in the pregnancy or in the labor would you do a rapid testing? What stage, if a woman presents to the maternity ward and she's in the first stage of labor, would you do rapid testing in order to give sufficient treatment?
MARGARET LAMPE: Sure, yeah. Actually, at any time -- it's really the sooner the better regardless of when it is when she presents. However, the data show and it's mostly observational data, quite honestly, that initiating prophylaxis just to the infant, even, within 12 hours of birth can still achieve rates of 13 percent transmission. So if she's in active labor and you don't want to talk to her about HIV testing or you can't get the sample or whatever, it doesn't mean abandon your efforts, it means as soon as she is done, send the sample down and get the test and you can still provide prophylaxis to the baby as soon as possible. And we are trying to get this on the minds of the triage nurse or the admission clerk or whomever, that this should be part of the standard labs. And we have worked with ACOG and we need to talk to the Hollister people, those are the people that create the form that many of the labor and delivery people work off of so that they have, you know, as you're doing your checklist, you say okay, they syphilis, did they have HIV too, or HIV also, excuse me. But yeah, as soon as possible, that's a good time. But I mean the ideal time is, again, as early in pregnancy as possible. Preconception is the best, but rarely does that happen. But really, if -- I mean doing rapid testing in labor and delivery is somewhat stressful for hospitals to do that, particularly to think about providing a women a positive result while she's still in labor, that's an anxiety provoking event. And so what we found somewhat anecdotally and in some places having this requirement, their prenatal testing rates go up because they want to avoid having to do it later and that's better for the women anyway to have it earlier and know it and be prepared one way or the other. I have a feeling we are running on time .
UNIDENTIFIED PERSON: We are okay.
MARGARET LAMPE: Anyone else? We have two great other presentations, too. Okay. Thank you.