DR. RICKETTS: I'm going to speak to you about how professionals are counted, give some examples, and discuss how that relates to problems in measuring access. Then I want to suggest some methods of analysis for use in policy making.

I think people are very familiar with this. It might seem very simple, but it turns out to be fairly complex in the world of working with data on health professionals and in dealing with policy questions. How do we count professionals? And, I'll focus on physicians throughout, with a few excursions into other professions.

Licensing-related inventories are probably one of the most common ways to count health care professionals. Most health professionals, by states or the jurisdictions, are defined, a process for licensing is prescribed, and a fee is charged. What happens after that varies from place to place. Some states create useful inventories based upon the process; other states do not.

There are other ways of inventorying and counting professionals. There are direct survey for inventory purposes. The AMA does it and states do it for various reasons for various health professions. That is, outside of the licensing process they survey the physicians to get an inventory. There are also indirect surveys that have been done for other projects. Indirect survey data could be looking at Medicaid data to determine who is doing deliveries.

There is no standard to this counting process. There's no way to say which of these methods are more superior in terms of the data they yield, because the data are treated so differently in each of the different ways that they are collected.

It would seem that licensing would be the best way to collect information on health professionals. But if that licensing is tied to a purpose other than data collection, you run into legal problems. That is, if the general assembly or legislature has set this up for a specific purpose and the reporting of information is not pursuant to the granting of that license, you might find yourself restricted in the use of that data. In other words, if you are using the data for purposes other than which they were intended to be used, and somehow it tends to constrain the market, then they will likely tell you that you can't use license data. It isn't common, but it is possible. Especially when you start talking about using distribution data to set up loan repayment for special programs that can affect a market. Then those data can become problematic if there's no clear intent on the part of the legislature or general assembly.

The Board of Medical Examiners, the Board of Nursing, the Board of Pediatric Examiners, and the Board of Optometric Examiners don't particularly care what the distribution of those providers are, according to their bylaws and the intents of those organizations. And they're not expected to do a good job in reporting it.

Someone has to identify the need for obtaining and summarizing these data, which makes direct inventory survey somewhat more usable in policy questions. That's why the state of Washington, for many years, licensed physicians but used direct surveys to find out where they were practicing. The state of Maine did the same thing for a while. Washington has since changed its process and put licensing into the hands of an analysis and policy group. Indirect surveys tend to be done for special projects, and are not always a useful source of information. There was a move in the 1970's to create a cooperative health statistic system under the sponsorship of the National Center for Health Statistics. North Carolina and 35 or so other states signed on because there was going to be Federal money to create standards for minimum data sets for health professionals, organizations, and institutions. Unfortunately, the funding was cut off rather quickly, and only a few states retained the function that was set up in the 1970's. North Carolina was one of those states. When the money was cut off, the states found that they couldn't afford to use these data, or to analyze them. And the reason was because the data were owned by the licensing agencies and someone had to be paid to take those data, get them out of the shoe boxes and off the 3x5 cards, put them into the computers, and report them. In North Carolina, it turned out that the area health education centers took this process on and the Sheps Center for Health Services Research cooperated with them. Since 1974, they have maintained the health professions data system out of their own budgets. The state does not pay any direct funding for inventorying health professions.

The physician license data comes as a physician applies for a license. You can ask them just about anything at the time of initial licensure, because the granting of a license is dependent upon a fairly loose definition of suitability to practice in the state. However, these data can only be collected one time, because it becomes a great burden in the re-licensing process to continue to collect all these data. So, if a provider goes on and does something later, or gets Board certification, there is a problem gathering these data later on.

I'm going to focus on the problems of practice location. When a person is initially licensed, in many instances, he/she is a graduate of the medical school in the state, and his/her address will soon change. Half of all newly licensed physicians that come into North Carolina change their address within two years. It settles down to about 30% within five years. But there's a great deal of change. In addition, there is no way to accurately understand the specialty of a physician given that the physician self-designates his/her specialty.

Board certification is one indicator. But suppose a physician is on the way toward Board certification, or has been Board certified in a particular area and decides to practice in another. In gathering re-licensure information, we only depend upon their honesty and accuracy. For example, we have been asking questions about whether physicians take Medicare or Medicaid patients, whether they do obstetrics, and whether they practice in an HMO or a clinic. In addition, people tend to put in a home or a mailing address, or P.O. boxes, which creates a big problem in tracking physicians.

Asking about activities, status, and specifics--such as whether people do deliveries--touches a nerve among physicians and other practitioners. They seem to think it is an invasion of their privacy, unless it is state mandated or related to some true public policy.

What information do we consistently get? We get race and gender most of the time and the location most of the time. But things begin to drop off as we ask more questions about where and how people are practicing. About 83% of the physicians indicate how many hours a week they're practicing medicine. You find that the response rate tails off as you begin to ask other things, like practice setting.

We are very concerned about where physicians practice and whether we are identifying them correctly as practicing practitioners, and whether they should be included in inventories of primary care availability. That's why we ask whether they have a part-time or full-time practice.

And then we ask about whether they have primary, secondary, or other practice locations, and what type of practice is in the secondary or other location. We've given some sort of allocation of time to these practice locations.

We asked the question: "Roughly what percent of your patients are prepaid/capitated, HMO, PPO?" in 1992, and received just a howl of complaints. There was some problem with individuals understanding this type of question. So, asking how many patients are in an HMO plan, if the physician works in a large medical school or some other setting, is difficult. If they work at Kaiser, it's pretty easy.

In North Carolina, based on the information we have, the primary practice locations are clustered around major cities. The secondary locations are spread more evenly across the state. The location of practices may be a lot different if you also look at where people practice secondarily. We haven't weighted this by hours. The secondary location could be screening in a school once a month. So, we'll look at that.

Now, I'm going to switch gears a little bit, because we want to talk about access. But we're wondering if those places that we have identified as underserved, and those places that look sort of like that but are well served, the question we were asking ourselves was: How can we look and see if we're -- how can we change this to make them adequately served? We wanted to determine the appropriate mix of providers for both well served and underserved areas. Basically, we were trying to figure out what was adequate care for a community.

We contrasted areas that had been adequately served by all measures, and those that had been chronic underserved areas, to see the difference. We tracked them over time. We did it for everything we had a measure of. We saw very little difference in the distribution of professionals, other than primary care professionals, except in the case of the R.N./L.P.N. ratio. There were more L.P.N.'s to R.N.'s in the chronically underserved areas than there were in the adequately served areas.

But we found a very interesting phenomenon. It didn't tell us anything about adequacy and inadequacy, but it showed us that you can change the provider mix and things won't really look different. In this case, two complementary groups, nurse practitioners and P.A.'s, switched places. The nurse practitioners left chronically underserved areas in a fairly steady stream until this past year. And, the P.A.'s moved into the underserved areas.

We still hold out the promise that universal coverage will allow us to track all these things and understand them. To do so, you're going to have to link files in the system. The material for creating linked files is not ready for prime time yet in most states, except perhaps New York, Maryland, and California. For example, UPIN numbers are not universally recognized. Even today, the accuracy of Universal Provider Information Numbers ranges from about 40% to 85%.

The cross analysis of files requires coding discipline. There has to be a real determination that coding will be excellent in terms of these index and indicator numbers. And you need four universal coding systems to link all this stuff in a useful way. I don't think we've seen this type of commitment.

JIM BERNSTEIN: I am going to talk about the relationship between an office like ours, which is the Office of Rural Health in North Carolina, and Health Services Research Center. I became very interested in this relationship when I was the Director of Union Health Service for Northern New Mexico and Colorado. There were a lot of individuals from public health schools who did studies. They'd come in and present papers. And in all my tenure there, I never made a change based on any of those papers. It wasn't necessarily the fault of the researchers, it was a fault of both of us. I didn't understand their language, or how to interpret it, or how to use it in a practical way. And they didn't know how to articulate to me what they were trying to say. And so, I did spend three years at the Health Services Research Center before I went to the state and established our Office of Rural Health.

We are interested in bringing primary care providers to underserved rural and indigent areas, and keeping them there. In our state, we have set up programs and policies to affect that. Loan repayments are one. Bonus payments another. We build health centers and we provide financing for them, too. And we have a number of field staff who are out every day working with communities to develop different programs.

But we need to understand what we should be doing and what we shouldn't be doing; where we should be spending the money; and what we should be asking the legislature for every year. And a lot of that comes from the interaction between the Research Center and our office. We sit down with their researchers and we say, "This is what we're seeing out there". And they'll take it a step further, and provide us with some funding to do in-depth interviews with physicians in underserved areas leaving the state.

We might do the same thing with physicians coming to the state. These are two- to four-hour interviews with physician and spouse. It gives us a lot of insight into the issue. And we think that we help to influence the questions that they ask, so they're more targeted and more pertinent to 1994 than they might have been in 1980. We have learned a lot from their surveys, and I think they've learned from us. It's important for us to be able to take their information, to understand it, and then to boil it down to something that can be used by our administration or legislature to create programs and policy.

We keep in front of our legislative committees the point that we are a rural state. We never let them forget that half the population live in small communities. We never let them forget that the resources--the physicians--that two-thirds of them are going to the one-half of the population that lives in the 17 counties and the rest of physicians, the one-third, are scattered among the 83 other counties.

I wanted to say something about the importance of the data that you present to the policy makers. More and more, the data you're putting together is affecting policies and programs. When you start putting your data together, at the end of that tunnel you're creating winners and losers. Some places are going to be winners, and some are going to be losers. And if your data aren't put together in a really thoughtful way, you could do a lot of harm.

MR. MARTIN: Good morning to all of you. I want to ask you to imagine that it is Monday morning. Imagine that you are the chief executive officer of a hospital, and you've just ended the Monday morning management team meeting. You're walking down the hall and what you have with you is a list of the problems that people around the table raised during the course of this week's discussion. The individuals at the meeting included the CFO. Typically, the CFO in your hospital is the first person to talk. The CFO says the financial ratios for your hospital are not too bad right now, although cash is in short supply, and you need to be a bit careful about that.

He goes on to tell you that the last round of layoffs that your hospital experienced are not yet producing any sort of cash flow benefits. There is also a friendly reminder from the CFO that the bond issue that you executed, in order to finance the ambulatory care center, requires that the hospital maintain certain financial ratios. And as if that's not enough, to conclude, the CFO tells you that the inpatient census is still falling faster than anticipated this year, and he's still not exactly sure why that's happening.

Next, the public affairs director tells you that Channel 5 News called this morning and they would like to interview you to talk about the sort of things that are happening in your community, in particular with regard to the glass factory that just closed.

At that point, the director of business development speaks up to say that the glass factory closing really has the business community spooked. And, in fact, the business round table in the community is beginning to talk very seriously about demanding from the local health insurers a five percent roll-back in premiums, if they even want to talk about doing business with them.

In addition, the business development director points out that the local Blue Cross plan is in serious negotiations to purchase a primary care practice, which happens to include a number of the physicians on staff at your hospital.

The professional services director then tells you that the schools in the community are also feeling the economic distress. They're concerned about the loss of tax base with the factory closing. And, they're beginning to hear that the school-based primary care clinic that your hospital staffs may be in jeopardy unless the hospital is able to come up with some of the funding that the school might not be able to provide.

The ambulatory care director mentions that a 50-year old woman who suffered a stroke came to the hospital emergency room. It turned out during the interview process that she's diabetic and that she had spent about four years managing diabetes herself without seeing her physician. Her recovery, at this point, is greatly complicated by the fact that she is massively depressed, not only because the consequences of stroke and what that's done to her capabilities, but also because she works for a dry cleaner and they don't offer health insurance.

The HMO director said that she had read over the weekend about an HMO that specializes in serving Medicaid patients. And they not only seem to be providing very good care and very good service, but they're financially viable. Can we get some resources of the hospital invested in exploring the possibility of a Medicaid HMO?

Now there are probably 30 or 40 problems on the problem list at this point. But a couple of things that you also noticed were an announcement from the American Hospital Association about the availability of the seminar on measuring access to ambulatory care and primary care. And you found below that a bill from the American Hospital Association for your annual dues, equivalent to one physician in the institution. As you look at that, you're thinking about the school health clinic and the problems there.

There's a lot more. Your mind begins to think about whether there is some common thread that would enable you to find a solution for any of these problems. And if it's common enough to cross problems, maybe you can solve a number of problems in one fell swoop.

Unfortunately, you don't have a lot of time to think about that, because you don't want to be late for a meeting to discuss the prospect of your hospital joining a network of providers forming in the community. That meeting also has to be squeezed in with another meeting with some of the hospital attorneys regarding a merger proposal from a different hospital.

Come back to today, and let me just offer a few points. As providers, the data community, the research community, and the policy community talk about measuring access to care, it's that sort of backdrop that will be in their minds. Every one of the issues that I have raised are real. Probably all of them you wouldn't find on any given Monday, in one particular hospital, but it's possible.

My expectation would be that in approaching providers with regard to their role in the study of access to communities they are going to be exceedingly impatient with any sort of analytic proposal that seems to set out to tell them what they already know. Providers believe they have a very good sense of the access problems that exist in the communities.

I think there's also likely to be skepticism if the initiatives are couched in terms of helping to build the case for new resources for improved access to care. My sense is, on a personal level, that if the job is to make the case for new initiatives in the community or new initiatives in the state, don't overrate it in order to sell it. I think it's more likely that the skills of a novelist, the skills of an anthropologist, the skills of someone who is able to describe in very human terms what these problems actually mean for a wide range of individuals is going to be much more compelling than statistics could possibly be at that point.

My guess is that where the most activity is going to be found within the provider community would be under two scenarios: One, if there's an organized health plan serving a defined population, that health plan is going to have a natural interest in understanding access care and in understanding the kinds of tools and techniques needed to identify the problems.

Second, where there are no organized plans, or there are segments of the population that are not served by organized plans, in that case, I think the best chance for involvement in those settings would be among a coalition that is concerned about the access problem.

These represent just a few quick comments on the provider perspective.