AMCHP 2005 ANNUAL CONFERENCE
DELIVERING RESULTS, IMPROVING PREGNANCY & BIRTH
February 19-23, 2005
DR. EMMETT GONZALEZ: Good morning from me as well. Obstetrics as most of know is a happy specialty. And it is filled with optimism. However, we do have those episodes that seem to crush at times and despite that we continue as Dr. Lu had stated earlier, we need to keep our optimism. So if it sounds a little depressing from some of the things that I may see, lets keep out optimism, because we do have a beginning. As Dr. Shabeer said I was in Eastern Washington working with a small hospital, city was about 30,000 and we were developing a perinatal program. It was really an improvement project, they already had an obstetrical unit. It is there that I want to begin with this story. Lucetta is a young prima gravita that was expertly delivered of her first pregnancy by a board certified obstetrician.
This is in Eastern Washington . It was great to have a board certified physician there to be able to accomplish this delivery. The delivery however, had to be by midforceps. I was there to witness the delivery. I happened to be in labor and delivery and he called me into the delivery room, actually was in a room, not the whole classical delivery room with all these big, you know, ugly stirrups. The torture rooms they call them. It was a regular room and when I looked at the baby after the deliver of the midforceps, the baby was absolutely beautiful, healthy, and didn’t have any forcep marks, which was absolutely remarkable. I marveled at this. About a week later, after the delivery, Lucetta died at a Seattle hospital. It was at a large medical center. She had died from overwhelming infection and with the diagnosis of acute respiratory distress syndrome.
Lucetta had total organ system failure, her lungs failed, her kidneys failed, her heart failed. Now, who is to say that Lucetta’s death was preventable and that the midforceps really create a circumstance in which there was a medical terminal event that was being setup, setup for this infection, this morbidity that Dr. Gould has eluded to in some of the elements of CPQCC. If Lucetta had been transferred earlier to the Tertiary Care Center in Seattle , could her life have been saved? And, what are the maternal quality indicators in this case. Now there is a definition that we use and there are a number of them really in terms of defining maternal mortality in the time frame. I personally had a patient who was a clinical psychologist. Her husband was an attorney. Her first baby was delivered by cesarean section. Two and a half years later, she wanted another cesarean section. She did have another cesarean section. Everything went well. Postpartum visit about six to eight weeks later, she wasn’t feeling well, and I did all the workups that most Obstetricians do for patients that come in not feeling well.
Talking about temps and so on and the training UC Davis is pretty powerful in internal medicine and I loved it and I spent six months of internal medicine as part of my internship. When she told me about some of the signs and symptoms, I thought maybe she had Hodgkin’s Disease. So I referred her to an internist friend of mine who did a thorough workup and she did have a lymphoma. Sandy died within the year of her second cesarean. Now the definition of that one be an associated maternal death. It wasn’t directly related to the Obstetrical care as in Lucetta’s death in Washington , state of Washington . These stories that I’m telling you are tragic. But we have to look at how is it that we can start gaining quite a bit from these tragedies. Now these tragedies may be more personal for me, however, there are occurring. They’re occurring every day and if we look at the numbers over the year, they are staggering for us.
Now there was a lead article in the public health journal this month a small one, nonetheless, it mentioned about public policy and maternal mortality. And it had a case example that was used and that was a case in India . So, I think that we need to look at improving this part. The California Maternal and Child Health has undertaken the development of a surveillance system to be able to establish a panel for maternal mortality. We don’t have one. In many states, we’re not alone, many state don’t have maternal mortality panels. The ultimate goal is to develop a public health with a system with partners that are in the healthcare field and to be able to reduce the maternal mortality and to have a goal as we need to have one to the HP 2010, the 3.3 per 100,000. The most glaring element about what I’m going to show you today and what has been previously shown in previous years is that the African American still suffers disproportionately in terms of maternal death.
This too, we need to pay attention to, very special attention to, and more recent work that was presented at the annual APHA meeting in San Francisco . One of the EIS officers whose now during her PHD at John Hopkins presented some data that shows that California Latinas are now appearing to be at higher risk for maternal mortality. I’ll show you who these women are that are delivering that are at higher risk in these different groups. The methods that we employ have to deal with maternal quality indicators and primarily because morbidity usually can start associating with mortality. So we want to tie those together. And the way we tie them is by utilizing the CPQCC model, that is looking at maternal quality indicators and how we can tie those to the associated maternal mortality figures.
So we’re going to employ that method of surveillance using the quality indicators and what I want to do in a very brief manner and we’ll jet along here a little bit, is the presentation of the demographics first of women giving birth in California. Then data that is concerning the maternal quality, maternal mortality rather of the population at a given birth, and subsequently give a description of a surveillance system that we anticipate that we’re going to put together on maternal mortality. It has to be on a panel driven basis, again using the CPQCC model and the maternal quality indicators. Okay, here we go, the first part is that saint motherhood has been incorporated in so many different ways and so many different countries, but our main emphasis is we want to use the maternal morbidity and tie that into maternal mortality.
We want to use the pregnancy related mortality review using the maternal quality indicators that the Southern California group has produced and will keep producing for us and CPQCC, Dr. Shabeer already eluded to that in regards to the quality indicators and as a matter of fact, two articles are coming out, three articles are coming out along this basis, not necessarily in terms of mortality but in terms of medical quality indicators. And this should be coming out this year. If we look at California births and in order to look at the big pie, let me see, well, I’m lasering myself. Hispanics make up the bigger part of this pie. In 2003 540,000 plus births and Hispanics made up as Dr. Shabeer pointed out on a different slide, 49.9 percent. This is white, non-Hispanic and it’s very important to look at the small fraction here of 5.7 on the African American and Asian of 7.2. If we look at from 1991 to 2003, by the way the 2003 figures just came out last Monday.
We didn’t get a chance to have some of the more high powered people do analysis on this with SAS and have more breakdown, but at least we got this and it is not very good, not very encouraging. But if you look at this, you see the curve going up, the trend, we had a 10.6 per 100,000 and for 2003 we’ve got a 15.2. Now if you were to super impose the U.S. mortality for the same number of years, ’91 to 2003, you would look very similar. The numbers may vary but certainly the curves and the downs and the ups would look the same. Now some causes of death? We’ll look at, the majority of them are going to be other direct obstetrical factors and this is a very important one, the eclamcia and preclamps. Now these are hypertensive diseases that can be addressed and should be addressed and we need a lot of partners for that. We need the clinical folks to be with us on all of this. And obstetrical embolism, right here, is this 10.9. Those are generally clots that have migrated to the lungs and cause the maternal death. The one element that I eluded to and the glaring element the disparities is right here.
These are the live births on this side. The African Americans only making up six percent, by the way, this is for 2003. And these are the maternal deaths. They make up 15 percent. If you look at 2002 figures, they made up 18 percent and the numbers are pretty much the same. They are like six percent of the live births at about 529,000 and they were 18 percent and they were six percent of the 529,000 for 2002. So really, we have no, we have made no change in fact maybe it’s gotten worse in terms of the absolute numbers. The work that was presented at the APHA, this is part of the work and what I wanted to emphasize were a couple of things. One is Hispanic women are still very much in the reproductive age years. Twenty-five years of age, that’s a median age, all women are 33. The other part that I wanted to really emphasize to you as well is foreign born women, this is form 2002, represent 32 percent of all California births. This is going to be important as you’ll see in a few minutes, in terms of the maternal mortality, the difference in foreign born women and 64 percent of all Hispanic births or from foreign born--
Emmett Gonzalez: And White Non-Hispanic and Hispanic are clustered down here. It’s still not down to the 2.3 per 100,000. Nonetheless, you can see a little blip here that will be more evident to you in a minute. This is the one where it shows the foreign-born and the US-born Hispanic. You can immediately tell who’s who. For the foreign-born women, their rate is 11.--is that 11.7 up there? 11.7, US-born Hispanic 6.8. Very dramatic, very dramatic and I think we’re not very conscious of that, because again, we’re still thinking about the paradox of the newborns. Despite the fact that all these figures may have a lot of limitations, because they’re small in number, sometimes the rolling average, they’re underestimated for a lot of reasons, and as you’ll see in a minute, there’s a lack of standardization sometimes in definitions.
So in definitions, who, for example, stays with the death of a pregnant woman within 42 days termination of pregnancy? They still use the direct and indirect in terms of cause. HB2010 uses the number of female deaths and the denominator as number of live births and again, this is the 42 days after pregnancy, it doesn’t go into the year, which CDC and Prevention and ACOG had agreed to. HB2020 again, the 3.3 per 100,000. ACOG and CDC and Prevention had to come up with the uniformity that we could all basically agree to. Well, most people could agree to. They divided into pregnancy-associated death, which is death of a pregnant woman within a year of termination of pregnancy respective of cause, and then the pregnancy-related death--now, one thing that I wanted to--of course, everybody wants to tinker with things, you know, I wanted to put a time associated, but that didn’t fly. Pregnancy-associated death is again the Lucero kind of concept, it was direct consequence of obstetrical care, and I won’t belabor going through all of this, but it’s the consequence of obstetrical care, its management, but not from accidental or incidental causes.
By the way, some of you have probably have had or have been acquainted with the work that was done in Maryland in which some domestic violence issues have now come up as very important in looking at who’s at risk for maternal mortality. And of course after those--that paper came out, there were a series of reports from the Washington Post that really dramatized this for the population. The pregnancy-related and -associated mortality in California , basically the players are--are MCH, A Branch, University of California , San Francisco and the Public Health Institute. The purpose, as has been with many of the programs that address this issue, is look at the medical psychosocial that led up to the death of a woman in a pregnancy-related or pregnancy-associated, the causes and to renew the interest in this public health problem. It is my contention that we don’t hear of maternal death very often and that’s why it stays in the background, but we really need to intensify our efforts, because this is going to be I think even a bigger problem when we start accepting the definitions that CDC and Prevention and ACOG has put forth, even though they’ve been out there for a number of years.
The way that we’re going to be doing it is, we’re going to be able to link birth file, fetal death file, the death file, the National Center for Health Stats and patient discharge data file. We needed all these files to be able to not miss or at least make the very best effort of not to miss any of the women that have died. The identified cases using these files will then undergo medical record review abstraction, a physician and trained labor and delivery--experienced labor and delivery nurses will look at the medical record. You know, when you’re going to do this at a state level, for those of you who may have a slight interest or even if you don’t, I hope that you support whoever may have an interest in your jurisdictions. It takes an enormous amount of effort to go through IRBs, Institutional Review Boards. We had to go through Institutional Review Boards at the UCSF as well as the state. UCSF is an academic institution and they understand a little bit more.
The state, unh-unh, even if you work with them and for them. It’s--it’s tough. There a lot of agencies, a lot of people that you have to speak to before you can get this done and get IRB Board approval. Although, I have to say that in this case, we have very good--very good support. It wasn’t as insurmountable. Of course I had Amy Godecker, who’s a Ph.D. that’s working with me, and she’s absolutely fantastic in drafting a lot of this material. Then after the medical record review, a case review team will determine where the case falls into. Specific questions that we want, how many women in California died that were not identified by the previous codes or the current codes? What are their primary causes and I think particularly we are interested if there are, at the bottom down here, are there any unique factors that contribute to this racial ethnic disparity? We know from the elements that not only were discussed this morning, but in previous presentations from other speakers and in the literature, the social determinants of health and all of these elements are important, but are there some specific things that we need to look at?
Our abstraction form for example took the Florida model, we added some spice from CDC and Prevention questionnaire, and then we looks at the UK abstraction form for their maternal deaths and put some of that in so that we could see if we could pick up some of the other social elements that we need to find out about for very specific areas of our population. And now we get into how we’re going to tie this to what the maternal quality indicators are like and what CBQCC is doing. And Dr. Gould already covered a lot of that and of course The Maternal Quality Indicator Group, as I mentioned, Dr. Corse, Dr. Kimberly Gregory, Dr. Michael Lou and others, this is what we’re doing. We’re looking at applying the model to maternal health. Now, in order to do that, and I think Dr. Gould may have alluded to some of this, is that we have to have indicators that almost everybody can agree to and everybody can see that they’re valid indicators. It’d be rate based that we can really have objective as to the care that the patient got, we have to have the ability to get good data collection and the cost of monitoring what we’re doing, has to be improvement potential as he mentioned in some of the elements about you have to have a way of improving what has been done, clinically and otherwise. And everybody has to know what happens, it can’t stay--whatever the results are, it can’t stay within the medical community or public health community, it has to go out to policy makers as well. And of course it is socially important.
Now, there are two that are very concrete and I wanted to make these concrete, because they’re easier to understand. Maternal quality indicators, and these are two examples, one of them, for example, is the perinatal infection secondary to labor and delivery management and the other one is postpartum hemorrhage. I’m going to take this one, postpartum hemorrhage, because that’s very important. And here’s the background, 28 percent of all US maternal deaths, complications including, you know, the hypervolemic shock, disseminated intravascular coagulation, renal failure, hepatic failure, adult respiratory distress syndrome. This is what Lucero had. And what can we--potentially preventable causes, okay? Traumatic deliveries, infections associated with uterine apnea leading to postpartum hemorrhage. And the objectives of the postpartum hemorrhage indicator--this is going to be, by the way, part of one of the published articles. I didn’t get all of the information from Dr. Corse on her--I did speak with her a few weeks ago and she was very willing to share all of her Power Point slides that she has, but it would’ve been a total new session on the presentation, because she has a number of indicators that she looked at. In any event, the main idea in her article that will be coming out and et al, is it describes postpartum hemorrhage appears to be associated with different hospital organizational structures, so you can take a look at an HMO, a community hospital, a district hospital, a university hospital and you can look at these and see how they are performing.
Also, within the hospital you can determine whether postpartum hemorrhage rates have a potential to be relevant to the hospital level. In other words, does the medical center tertiary care center and a community center level or let’s say, let me put it to you this way, a medical center level tertiary care center and another one in another city tertiary care center, do they have similar rates. You have to compare those. You have to compare medical centers to medical centers, community hospitals to community hospitals, etcetera, etcetera. That’s how we want to be able to get these indicators, to be able to look not only at the type of hospital, but also at its peer hospital if you want. Now, what they did is that they took OSHP data, ’97, that’s the Office of Statewide Health Planning, and they took all laboring mothers, term singleton, they excluded neonates with congenital anomalies and malignancies and excluded moms that have maternal fetal or placental complications.
They made some exceptions with the patients who had some choryeminitis or prolonged rupture membranes. They also had very good case definition, whether the hemorrhage occurred immediately, delayed hemorrhage, stage III, and whether hemorrhage was secondary to a coagulopathy, a bleeding disorder. Also, whether there was trauma to the perineum and vulva during the delivery and high vaginal cervical lacerations or pelvic hematomas. So using this they were able to get a lot of data, a good presentation of how we can approach this problem of indicators, they’re associated with morbidity, morbidity associated with mortality, and we can take a look at that. So what do we want to do? Keep having a better data driven system; as Dr. Gould alluded to, you need data. Keep using the CBQCC, public health programs, professional and clinic organization that can help us, ACOG’s on board already with this review that we’re going to be doing. And get a statewide surveillance system that can be established and sustained.
The upcoming trends and by the way, I need to make a remark about this, that the pregnancy check box is part of the death certificate now is there as of 02-03, and as I was leaving, one of our epidemiologists was telling me that because of being able to check on the box on the death certificate whether pregnancy had occurred within the year, the numbers, whether it be associated or pregnancy-related, are going to be higher. They’re going to be higher, which means, again, as we all suspected, we’d been underreporting. Also establish a near miss morbidities and a means of monitoring those, accessing the outcomes for the severity, create--and this is a very good one if we can do this, this create a maternal birth cohort file which would include all births linked to pregnancy-related hospitalization and deaths occurring within one year postpartum. That’s going to be a task. Dr. Shabbier, can we do that? He’s my chief, I’ll put him in the spot here. And we need active surveillance of all pregnancy-associated morbidity and mortality, continue working with the CBQCC and P (inaudible) model.
I’m very excited about the elements of what the panel does, because I have been in some of those meetings for the panel and the exhaustive review of the literature that’s done before the project is undertaken is enormous. And use the maternal quality indicators as I mentioned and the establishment of a statewide maternal mortality review panel that will be permanent. We anticipate, and this is what we want and we hope that we can get a lot of support from everyone within DHS and from outside clinical organizations as well. Thank you.