Printer Friendly

Reaction 2.

Pamela Farley Short, Ph.D., Center for Health Policy Research, Pennsylvania State University

At one point, we had a very simple idea of access in mind. But as we go along, we expand that notion. We have to decide where we want to stop expanding the concept of access so that it remains meaningful. If you look at some of the materials just reviewed, one could easily take out the word "access" and insert "efficiency and equity of health services" or "system performance." We have to decide if we want access to mean something special.

I would say a really simple definition of access is "getting in," particularly in the context of contrasting people who are uninsured, or severely disadvantaged, with the rest of us. "Getting in," having any visits at all, unmet need, waiting time, travel time, do you have a usual source of care - really rudimentary things - have been measured very satisfactorily in population surveys. But as we go along, we want to look at "getting in deeper," not just getting in the door, not just the uninsured versus the rest of us, but also at people who have different health insurance arrangements. Then our focus is no longer just on initial access, but on people as they move deeper into the system, access to specialists, and access by people with chronic illnesses. This is far more complicated.

When we were thinking about just getting in, then population-based surveys were almost the only way to go. Any other approach misses the uninsured. You have to start with the whole population because, by definition, uninsured people are not in the insurance system and often are not in the healthcare system. If you want to measure "getting in," you have to know about the people who remain "out."

As we go into the system deeper and our questions change, then we need to think more carefully about whether or not a population-based, random-digit-dial, go-and-select-some-doors-to-knock-on approach is really the way to go, or whether there are better sampling approaches.

So I see that we are expanding access as a concept in two ways. The first is along one dimension by following people deeper into the system. The second is by wanting to measure "access to quality care," or other modifiers, like "appropriate," "the right," "effective," and "efficient" care. This dimension clearly has to do with normative value judgments. These are not objective measures; they imply notions of good and right.

In this context, the easiest thing to do is to look at equality as our measure of good and right. Is everyone getting the same thing? Equality is easy to measure but it is not very satisfying; it really doesn't fit our idea of fair and equitable. Just ask anyone who wants to risk-adjust some performance measure. The same is not necessarily the same.

As we push further along normative lines, the first place we come to, after inequality, is judgments of "adequacy" or "minimal acceptability." Here we may want to include poor clinical outcomes; survey questions that probe for unmet needs, something you really needed - not just something you wanted and you thought would be nice; and serious difficulty paying for care that patients did get. This line of inquiry challenges us to arrive at clinical and social agreement on where to draw the line in the sand, where we are unwilling for anyone to go past this line.

I think we have made some pretty good methodological progress in this area. Some survey questions resonate with us. The answers make us squirm in our seats and - if we are talking about minimally acceptable and adequate versus inadequate - then we are probably at the right place.

Next, in assessing access to good care we begin to make trade-offs against costs and also let personal preferences and different values enter. As we move in that direction the demands on our measures and for data become more and more difficult. An immediate problem we encounter is . . . is it fair? Well, it depends. It depends on the circumstances. It depends on the kind of health problems someone had and whether or not they really needed to see the doctor. Expanding the definition of access this way pushes our data collection to focus more and more on specific circumstances and subgroups. And more specific circumstances and subgroups require larger samples. For example, in order to look at access to specialists, the target population is people who needed or used a specialist within the reference period. This is only one-third of a sample of the general population of adults, from what we're finding in the Consumer Assessment of Health Plans Study (CAHPS). So it is necessary to triple sample sizes to look at this question.

Normative measures require careful clinical judgments that, in turn, require more information. Here the problem with population-based surveys is not only limits on sample size but also people's ability to actually tell you the nuances of their healthcare problem in a way that allows the researcher to make valid judgments about their care.

This brings us to linkages. Some linkages are already in place. We can now study more specific subgroups of people with chronic illnesses or severe health problems by identifying them in the National Health Interview Survey and piggybacking other surveys onto it. Thus, one big survey - NHIS - can be used to find a sample for a smaller survey.

The limits on what we can ask household respondents also push us to link population-based surveys with administrative records and enrollment files to identify people with different insurance arrangements. In fact, it is probably easier to measure access than to measure what we mean by different types of health insurance. There are two possible ways to solve this problem: (1) interview household respondents and then go to more knowledgeable respondents, the insurer and the employer, to get the necessary insurance information, or (2) start with enrollment files to draw a representative sample of people with the kind of insurance that we want to study.

The Consumer Assessment of Health Plans Survey (CAHPS) has taken the progression on insurance typologies to its last stage. We started with the uninsured versus the insured. Then researchers were struggling with, What is an HMO, what are plan types? Jack Hadley has proposed pushing a bit further to focus on certain salient characteristics of health plans. CAHPS and a number of other large payers say "forget all that, I want to know about specific plans and will go directly to their enrollment files." This way we know all about the plan in terms of coverage, benefits, and other characteristics. We can ask consumers what their experiences are and measure all kinds of things. This is the way we may find ourselves going more and more in the future.
COPYRIGHT 1998 Health Research and Educational Trust
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:response to Judith D. Kasper, in this issue, p. 715, and Raymond Fink, in this issue, p. 741
Author:Sennett, Cary
Publication:Health Services Research
Date:Aug 1, 1998
Previous Article:Reaction 1.
Next Article:Reaction 3.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters