| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Sociology, George Washington University, 2121 Eye St., NW,
Washington, DC 20052.
2 Research Department, American College of Radiology, 1891 Preston White Dr.,
Reston, VA 20191-4397.
3 Department of Diagnostic Radiology, Yale University, New Haven, CT
06520.
Received May 12, 2003;
accepted after revision July 23, 2003.
Address correspondence to J. Sunshine
(jonathans{at}acr.org).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. A survey was mailed in 1999 to a sample of 970 radiology practices; completed, usable surveys were returned by 66% of the practices. Three indicators of managed care were used: a practice's percentage of managed care (HMOs plus preferred provider organizations), local area HMO penetration rate, and self-reported perceived effect of managed care.
RESULTS. Percentage of managed care averaged 30% but was 40% for multispecialty groups. It was relatively high in large metropolitan areas, for practices with no hospital activity, and for practices with any owners who were not practice members. The three measures of managed care were only moderately correlated (correlation coefficient, 0.25–0.33). None of the managed care variables had a statistically significant effect on days provided for vacation and continuing medical education, promptness of payment, years required for practice ownership (partnership), and percentage of practice members who were owners. Higher percentage of managed care was associated with higher collection rates, whereas greater perceived impact of managed care had the opposite association. Two thirds of practices belonged to at least one managed care–related organization such as an independent practice association. Most radiology practices reported no involvement in the managed care negotiations of hospitals, which was true even when the hospital's negotiations included the radiologists' fees or when the practice determined its level of involvement.
CONCLUSION. Many negative outcomes most feared by radiologists regarding the effect of managed care have not materialized. Perceptions of practices as to the effect of managed care seem to reflect negative aspects of their general situation, not only realities of managed care.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The decade of the 1990s witnessed a rapid growth in managed care, along with many other changes in the professional and business aspects of radiology practices [2–4]. Research based on the ACR 1995 Survey of Radiologists reported a negative association between self-perceived (but not actual) level of practice involvement in managed care on the one hand and professional satisfaction of radiologists on the other hand [5, 6]. Research on the hiring patterns of radiology practices in 1996 and 1997 [7, 8] found that radiology groups experiencing the greatest self-perceived impact of managed care on the practice were the least likely to report expansion positions or to offer positions on a partnership-track basis. However, data on 1998 hiring collected as part of the ACR 1999 Survey of Practices showed that perceived impact of managed care had no effect on hiring plans in 1998 [9].
The ACR 1999 Survey of Practices provided an opportunity to examine the extent of involvement of radiology practices in managed care at the end of the decade, as well as the impact of managed care on radiology practices. Exact numbers, such as the average percentage of practice in managed care, change from year to year, if not from month to month. Therefore, in this article, using data from the 1999 survey, we concentrate on more important and longer term relationships—for example, whether the negative effect of perceived involvement in managed care has persisted while actual level of involvement has continued to show no negative effects.
First, we examine the relationship among three measures of managed care: actual percentage of practice in managed care, average managed care level for the local geographic area, and subjective perception of the impact of managed care on the practice. We then investigate how the level of involvement in managed care varies with different types of practices, such as large versus small practices, academic or multispecialty groups compared with other private radiology practices, and in different geographic regions of the country. After analyzing factors that affect practices' percentage of managed care and the perceived impact of managed care, we then ask whether, and to what extent, the level of managed care involvement affects other aspects of practice life such as days off for vacation and continuing medical education (CME), net and gross collection rates, use of physician extenders, years required for ownership (partnership), membership in organizations such as an independent practice association, and relationships with hospitals.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The random sample of 970 practices was drawn from a list of 2,448 responses to a 1998 mailing by the ACR. A total of 708 responses were received. Of these, 30 responses were excluded for being out-of-scope (e.g., residents in training), another 33 were excluded for incomplete answers on critical questions such as the number of radiologists in the practice, and 28 were duplicates. The 617 usable, distinct responses from a possible 940 (970 – 30 out-of-scope) constituted a 66% response rate. The response rate was close to the 69% achieved on the previous (1992) ACR Survey of Practices [1]. As explained elsewhere [9], analysis of potential nonresponse bias was not possible.
For the data analysis, the usable responses were weighted to make them representative of the total population of radiology practices in the United States, which the ACR has estimated to be approximately 3,010 multiradiologist groups and 1,556 single-radiologist practices [10]. As described elsewhere [4], the usable responses were divided into seven practice sizes (measured by the total of full-time and part-time radiologists in the practice at the time of the survey) in each of the four census regions, for a total of 28 categories for weighting. Although practices that sent duplicate responses were counted only once in computing the response rate, as described in the preceding paragraph, both their responses were used in the data analysis, with each of the duplicate responses assigned half the weight otherwise appropriate.
For this article on managed care, the eight responses from government practices were omitted because government practices often do not deal with managed care or any other payers. (The government practices are included in most other papers based on the 1999 survey.) Thus, this paper is based on information from 609 practices.
In addition to means (averages), percentages, and other statistics from the survey, we present their SE, which is the usual quantification of their uncertainty.
Variables Defined
Three measures of managed care were used. One was the percentage of
activity of a radiology practice in managed care, based on data reported by
the practice. Practices were asked to indicate approximately what percentage
of the practice is in each of nine categories: HMO, preferred provider
organizations (PPO), Medicare, Medicaid, workers compensation, indemnity plan,
contract interpretation, self pay, and other. Although survey instructions
indicated that percentages should total 100% for the nine categories, not all
actually did. If the total for all categories was less than 80% or more than
125%, the data were not used because the error was judged to be too great. For
3% of responding practices, data on the percentage of practice in managed care
were deleted because of this error. For all other responses, we took the sum
of HMO plus PPO percentages divided by the total for the nine categories. We
refer to this measure as the percentage of managed care, which has a possible
range of 0–100%.
The second managed care measure was the HMO penetration rate in the practice's geographic area. This is the total HMO enrollment in a geographic area divided by the total population of that area, which has a possible range of 0–100%. This measure is a contextual variable describing the external managed care environment in which the practice operates. We used 1998 HMO penetration rates (i.e., data approximately contemporaneous with our survey) constructed by Douglas Wholey and other researchers at the University of Minnesota [11]. For practices located in metropolitan statistical areas (based on mailing zip code), data for the metropolitan statistical area level were used; and for practices outside a metropolitan statistical area, county level data were used.
As a third managed care measure, we examined a subjective indicator of the impact of managed care on the practice. The survey asked, "Overall, how much has managed care changed your practice to date?" We did not define managed care. We referred to this measure as the practice head's perception of the impact of managed care on the practice or, more simply, as perceived impact of managed care. Five response options were provided: none, little, moderately, significantly, and greatly. We coded the answers on a 5-point, equal-interval scale where "none" equals 0 and "greatly" (the highest category) equals 100.
Type of practice, coded into five categories, was an important variable used throughout the analysis. Any practices self-identified as primarily academic were categorized as academic practices. Among nonacademic practices, government practices such as Department of Veterans Affairs or military practices, but not private groups on contract to the government, were classified as government practices and are omitted from this paper. Nonacademic, nongovernment practices that were part of a multispecialty physician group were categorized as multispecialty practices. "Private radiology practice," by far the largest category, was reserved for practices that were nonacademic, nongovernment, and not part of a multispecialty group. Responses from practices that checked "other" were either placed into one of these four categories on the basis of additional information provided or, if they did not fit the categories, were relegated to a miscellaneous "other" designation.
For size of local area, six response categories were available, as shown in Table 1. For the regression analyses in Tables 2, 3, 4, the size of the local area was instead coded more concisely as a 4-point scale on which large metropolitan area is 4; small metropolitan area, 3; nonmetropolitan city or town, 2; and rural area, 1.
|
|
|
|
Definitions of most other variables are either self-evident or explained in the text that follows. The exception consists of two questions about collection rates. The survey asked for the gross collection rate, with instructions to calculate it as net revenue (total patient revenues minus refunds) divided by gross charges (total charges). The survey also asked for net collection rate, defined on the questionnaire as net revenue divided by net charges (defined as gross charges minus allowances such as managed care contracts and Blue Shield, Medicare, and Medicaid allowances, but not minus bad debt or charity).
Statistical Analysis Methods
The first part of the data analysis uses a practice's managed care
involvement as a dependent variable, or outcome to be explained, asking what
characteristics of radiology practices and external factors (such as local HMO
penetration rate) are associated with higher or lower involvement. The second
part of the analysis uses the practice's level of managed care involvement as
one of the independent (explanatory) variables, asking how managed care
affects other aspects of the practice such as collection rates. The final
section focuses specifically on the impact of managed care on contracts and
other relationships with hospitals. In all sections, we include both
descriptive statistics such as means, percentiles, and percentages, as well as
multivariate regression analyses. The latter enable us to sort out the
importance of each of several factors, holding constant the effect of other
relevant factors. In general, we take a two-tailed p value equal to
or less than 0.05 as the criterion of statistical significance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Percentage of Practice in Managed Care
Multiradiologist diagnostic-only practices averaged 31% of the practice in
managed care in 1999 (Table 1).
This figure included a range of 0–100%, with the lowest one quarter of
all groups having 17% or less in managed care, and the highest quarter having
43% or more in managed care.
The percentage of practice in managed care of multiradiologist diagnostic-only groups varied by characteristics of the group. Multispecialty groups showed high involvement, with an average of 40% of the practice in managed care (Table 1). Larger groups reported more managed care than smaller groups, with groups of two radiologists averaging 28% of the practice in managed care compared with groups of 15 or more radiologists averaging 37% in managed care. Groups that practiced in hospital settings, either exclusively or in combination with other settings, averaged 28–30% of the practice as managed care, whereas groups with no hospital practice at all averaged 47% managed care. Radiology groups that were entirely member-owned—that is, all owners were radiologists in the group (although not all group members were necessarily owners)—reported an average of 29% of the practice in managed care. In contrast, radiology groups completely owned by outsiders (which can include university-owned practices) averaged 43% of the practice in managed care, and groups with a mix of member and outside ownership averaged 46% of their practice in managed care.
The percentage of the practice that consisted of managed care also varied
with the geographic location of the radiology group. Groups located in large
metropolitan areas (population 
1 million) averaged 40–41% of the
practice in managed care. Regionally, managed care involvement was highest for
radiology groups in the West (37% of practice in managed care) and lowest for
groups in the South (27% of practice in managed care).
Two other types of practices were also studied. Solo radiology practices reported an average of 28% of the practice in managed care, slightly lower than the 31% for diagnostic multiradiologist groups. Mixed radiation oncology and diagnostic radiology groups reported 33% of the practice in managed care.
Regression analysis results are shown in Table 2. The numbers in the table are the regression coefficients—that is, the change in the dependent variable associated with a unit change in each explanatory variable after statistically controlling for the effects by all other explanatory variables considered. For example, the 0.32 regression coefficient for local area HMO penetration rate in column B means that, other characteristics being equal, the percentage of managed care was 0.32 percentage points higher for each additional percentage point of HMO penetration in the local area.
Column A of Table 2
indicates that most of the relationships observed in
Table 1 are statistically
significant (p 
0.05) even when effects of all of the other
factors are controlled for statistically. For example, using private radiology
group as the comparison (or reference) category, multispecialty groups had
more managed care (
6% more) and academic and all other type groups had
less ( 18% less for academic groups), all other characteristics being
equal. However, regional differences were not statistically significant, nor
were differences between mixed (diagnostic and oncology combined) groups
compared with diagnostic-only groups.
Column B of Table 2
introduces an additional explanatory factor—the HMO penetration rate for
the geographic area considered. If the practice characteristics that are
statistically significant in column A were no longer significant in column B,
this would indicate that local area HMO penetration, not the characteristics
of the practice, is the main explanation of percentage of managed care.
However, such was not the case. All the characteristics that were
statistically significant in the first regression equation (column A) remained
statistically significant (p 0.05) in column B, with the
exception of number of radiologists in the practice, and the size of their
effect remains similar.
Nonetheless, the HMO penetration rate in the area is clearly a statistically and substantively significant factor. The data indicate that for practices with similar characteristics for the other variables included (practice size, ownership, practice type, and so forth), a 10-percentage-point increase in the HMO penetration rate in the area was associated with approximately a 3-percentage-point increase in a radiology practice's managed care percentage.
Perceived Impact of Managed Care
In response to the subjective question about how much managed care has
changed the practice, 3% of respondents answered "none," 21%
answered "little," 36% answered "moderately," 28%
answered "significantly," and 12% answered "greatly."
In the regression analysis in column C of
Table 2, we investigated what
factors shaped the response to this question. Holding constant all other
characteristics considered, a higher percentage of managed care and a higher
local area HMO penetration rate were associated with greater perceived impact.
Other characteristics being equal (including the other two managed care
variables), private radiology practices perceived more change due to managed
care than did any of the other types of practices (academic, multispecialty,
or other). Other characteristics being equal (including the other two managed
care variables), practices in large metropolitan areas and in the South or
West perceived more change due to managed care. Practices with either
hospital-only or non–hospital-only settings perceived less impact than
otherwise identical practices with both hospital and nonhospital practice.
Practices with any outside owners reported more change related to managed care
than otherwise identical completely member-owned practices. Thus, a number of
the same variables that predicted percentage of practice in managed care also
predict the perceived impact of managed care above and beyond their effect on
actual percentage of managed care. (Their effect on percentage of managed care
is controlled by inclusion of that variable.)
Effect of Managed Care on Business and Professional Aspects of Practices
In Table 3 we ask, How do
levels of managed care involvement, as quantified by all three managed care
measures, affect other aspects of the practice? Other possible explanatory
factors, such as practice type, region, and ownership, are included as control
variables. Only multiradiologist practices are examined in
Table 3, and the table shows
the regression coefficient for only those effects that are at least marginally
statistically significant (p 0.1) and indicates whether an
effect was positive or negative and the level of statistical significance.
When no coefficient is shown (blank cell), the effect was not even marginally
statistically significant.
Table 3 shows that none of the three managed care measures had any statistically significant effect on the number of days provided for vacation and CME. All else being equal, multispecialty groups tend to report approximately 9 fewer days provided for vacation and CME than private radiology groups, and groups in the West report approximately 7 more days for vacation and CME than those in the Midwest (the reference category).
For multiradiologist groups in which at least some of the owners were radiologists in the practice, we investigated how managed care affects the path to becoming an owner or partner within the practice. In 1999, groups required an average of 2.9 years for ownership. On average, groups with any member owners reported that 78% of their member radiologists were practice owners, 16% were on an ownership track but not yet owners, and 5% were non–ownership track employees or contractors.
Table 3 shows that the managed care variables had no statistically significant impact on length of time required for ownership, no effect on percentage of member physicians who were owners, and no effect on percentage on an ownership track or on percentage not on an ownership track. Rather, differences in ownership path were affected by other factors. For example, other group characteristics being equal, groups with any outside owners had a smaller proportion of practice radiologists who were owners and a larger proportion on an ownership track but not yet owners. Interestingly, the proportion of non–ownership track radiologists (including contractors) was not significantly affected by the presence or absence of outside owners. Other group characteristics being equal, time required for ownership was highest in the Northeast, percentage who were owners was lowest in the South, percentage on an ownership track but not yet owners was highest in the South, and percentage on a non–ownership track was highest in large metropolitan areas.
We also investigated whether managed care affected collection rates or collection time (data not shown). The managed care measures affected gross and net collection rates but not collection time. Other characteristics being equal, a higher percentage of a practice involved in managed care was associated with a higher gross and net collection rate. The effects are statistically significant but small: a one-percentage-point increase in managed care was associated with a 0.08 percentage point (SE, 0.04) increase in gross collection rate and a 0.10-percentage-point (SE, 0.05) increase in net collection rate. The perceived impact of managed care had the opposite effect on collection rates. That is, greater perceived change in the practice due to managed care was associated with a lower gross and net collection rate, all else being equal.
Participation in Physician and Practice Networks
The survey asked whether practices belonged to any of the following types
of physician and practice organizations related to the development of managed
care: independent practice association, physician hospital organization,
physician organization, provider-sponsored network, physician practice
management company, group practice without walls, or other. As shown in
Table 4, overall, two thirds of
all radiology practices belonged to at least one such association, with 73% of
multiradiologist practices and 56% of solo practices belonging to at least one
of the organizational types listed on the survey.
Overall, an independent practice association was the most common association type, with 40% of all radiology practices belonging to an independent practice association, followed by 33% in a physician hospital organization, 17% in a physician organization, 7% in a physician-sponsored network, 4% in a physician practice management company, and only 1% in a group practice without walls. Among multiradiologist practices, 42% belonged to a physician hospital organization and 38% belonged to an independent practice association. In contrast, if a solo practice did belong to an association, by far it was most likely an independent practice association; 44% of solo practices belonged to an independent practice association. Twenty-nine percent of multiradiologist practices belonged to two or more of the organizational types, with 16% belonging to both an independent practice association and a physician hospital organization association.
We used logistic regression analysis to examine what factors affected whether a multiradiologist practice belonged to any association at all (Table 4). Other practice characteristics being equal, the managed care and geographic location variables were not statistically significant in predicting whether a practice belonged to some organization versus none at all. But the likelihood that a multiradiologist practice belonged to some type of managed care–related association was greater for large practices, for practices without outside owners, and for practices functioning in both hospital and nonhospital settings.
For practices belonging to at least one of the six association types, we investigated what practice characteristics predicted membership in an independent practice association or in a physician hospital organization. Results from logistic regression analysis (Table 4) indicate that, other factors being equal, a higher percentage of managed care at the practice level and a higher HMO penetration rate in the local area were both associated with increased likelihood of membership in an independent practice association. Academic and multispecialty practices were less likely to belong to an independent practice association or a physician hospital organization than were private radiology practices. Practices in the West were the most likely to belong to an independent practice association, whereas practices in the Midwest were the most likely to belong to a physician hospital organization. Practices that belonged to an independent practice association were less likely than others to belong to a physician hospital organization and vice versa.
Over and above the effect of region, we investigated whether practices in California, as sometimes speculated, were more likely to belong to an independent practice association or other organizations than were practices in other states. All else being equal, no statistically significant California effect was seen.
Use of Physician Extenders
Approximately 7% of practices reported using physician extenders (i.e., mid
level practitioners such as physician's assistants or nurse practitioners)
(Table 4). We found that among
multiradiologist practices, other practice characteristics being equal, the
use of physician extenders was not affected by percentage of managed care.
However, the greater the perceived impact of managed care, the less likely the
practice was to use physician extenders. Larger practices and multispecialty
practices were, other things being equal, more likely than others to employ
physician extenders.
Managed Care and Relationships of Radiology Practices with Hospitals
This section examines issues related to hospital practice arrangements that
may be affected by, or be a result of, managed care. One quarter of the
practices surveyed indicated that they did not function at a hospital at all,
and those practices were excluded from the analysis of hospital relationships.
For practices that functioned at multiple hospitals, the data are from a
single, randomly chosen hospital. Approximately 83% of radiology groups with
hospital practices indicated that the group is involved in the same health
plans and managed care organizations as the hospital. Less than 2% of
radiology groups with hospital practices reported that the hospital charged
the practice a percentage of the practice's professional fees.
Table 5 reports responses to survey questions about managed care contracts and hospitals. When asked about the typical role of the practice when the hospital negotiates a managed care contract, most practices (58%) reported that the hospital does not involve diagnostic radiologists or radiation oncologists at all in the contract negotiations. Although some practices (22%) indicated superficial participation, only 20% of practices reported more serious involvement in at least some contract negotiations (Table 5). An even smaller percentage—only 8%—of practices reported that the hospital seriously involves radiologists in "most or all" managed care contract negotiations (not shown in table).
|
When asked who determines the role of the practice when the hospital negotiates a managed care contract, 20% of practices answered that it was the hospital administration only, another 41% reported that it was the practice only, and 39% responded that both the hospital and the practice decided (Table 5). Table 5 shows that when the practice alone determines involvement in the contract negotiation, more than 68% of practices have no involvement. In other words, practices are more likely to have no involvement when they alone determine their level of involvement in negotiation than when the hospital participates in the determination. On the other hand, serious involvement of radiology practices with hospital managed care contract negotiations is most likely to occur (31%) when both the hospital administration and the practice determine whether the practice will be involved, and is least likely to occur (5%) when only the hospital administration decides.
When asked what fees the hospital negotiates in managed care contracts, most practices (77%) reported that the hospital negotiates only the technical component of managed care contracts. Table 5 also shows that practice involvement is lowest when only the technical component of the fee is negotiated. Specifically, for hospitals that negotiate only the technical component, 61% of practices report no involvement in contract negotiations. When the global fee is negotiated by the hospital (14% of practices), practice participation is slightly higher, with only 52% reporting no involvement, but the change is from no involvement to superficial involvement; serious involvement is no higher. Practice involvement is greatest for the small number of practices (10%) in which fees negotiated by the hospital were reported as "other"; only 38% of this group reported no involvement.
Table 5 shows that approximately 14% of practices were in a situation in which the hospital negotiated global fees. The same column of Table 5 also shows that for most (52%) of this 14%, the radiology practice had no involvement at all. Thus, in at least 7% of all cases (52% of the 14%), hospital contracts require practices to practice in managed care plans that the practices do not negotiate at all.
We used logistic regression (not shown in tables) to investigate what types
of multiradiologist practices were most likely to be in hospitals in which the
hospital negotiated something other than simply the technical fee—that
is, when the radiologists' fee was probably included in the payment hospitals
were negotiating. We used the same list of independent variables shown in
Table 3 (but without
"non–hospital only"). We found three practice
characteristics were, other things being equal, positively associated with
being in a hospital that negotiated something other than the technical fee:
having any outside owners, having a hospital-only practice, and location in a
large metropolitan area. In contrast, a location in the West or South, a large
number of radiologists in the practice, and a mixed (radiation oncology and
diagnostic radiology) practice were, other things being equal, associated with
being in a hospital that did not negotiate anything other than the technical
fee (p 0.05). Higher HMO penetration rate in the metropolitan
area had a marginally statistically significant (p < 0.1) negative
effect on negotiating anything other than the technical fee. The percentage of
managed care of the practice and the perceived impact of managed care had no
effect on what the hospital negotiated.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The subjectively perceived impact of managed care on the practice and the more objective indicators—percentage of managed care at the practice level or HMO penetration at the local area level—were only moderately linked. In fact, in some cases (gross and net collection rates), percentage of the practice in managed care and perceived impact of managed care on the practice showed opposite relationships.
Overall, we found little or no evidence to support many of the negative predictions and assessments often offered with respect to the impact of managed care on radiology practices. None of the managed care indicators had measurable effects on practice behavior in areas such as number of days provided for vacation and CME, years required for partnership, or managed care contract negotiations with hospitals. The lack of effect on time to partnership was, in particular, surprising. Similarly contrary to expectations, net and gross collection rates actually were higher when the percentage of managed care was higher. However, in contrast, net and gross collection rates were lower when the perceived impact of managed care was higher.
One interpretation of these findings for the perceived impact of managed care on collection rates is that when net collection rates are low, practices tend to attribute their misfortune to managed care even when the actual effect of managed care may not be harmful. (Indeed, as noted, the effect on collections of actual percentage of managed care appears to be positive.) To test this thesis that poor financial results lead to managed care being blamed, we reversed the analysis and considered net collection rate as a predictor of perceived impact of managed care (rather than perceived impact as a predictor of net collection rate, as in Table 3). We found that practices with a lower net collection rate but similar actual involvement in managed care and similar other characteristics, reported a larger perceived impact of managed care on the practice. Perhaps the perceived impact question taps dissatisfaction with managed care. We have seen a similar pattern previously, in a 1995 survey of individual radiologists. Lesser professional satisfaction was associated with a greater self-perceived effect of managed care but was not associated with any objective measures of managed care [5, 6].
Membership in Managed Care–Related Associations
In the analysis of membership in managed care–related organizations,
we investigated whether a California location had an effect above and beyond
being in the West. California is often credited with originating the physician
and practice networks and organizations developed in response to managed care.
Among multiradiologist practices, California practices were not more likely to
join a managed care–related organization, other practice characteristics
being equal. Among the practices that did belong to at least one such
organization, those throughout the West, not especially those in California,
were most likely to belong to an independent practice association.
In origin, the idea of an independent practice association was that it would be the single, all-purpose vehicle for practices' dealings with managed care. Our data suggest that as of 1999, that was far from the situation. Although solo practices were much more likely to join an independent practice association than some other type of association, multiradiologist practices were about equally—in fact, slightly more— likely to join a physician hospital organization as to join an independent practice association. Only about half (52%) of multiradiologist practices that belonged to an association for dealing with managed care belonged to an independent practice association. About half (53%) of multiradiologist practices that belonged to an independent practice association also belonged to another type of association. Of the multiradiologist practices that belonged to just one association, roughly equal numbers belonged to an independent practice association and to a physician hospital organization. Practices that belonged to a physician hospital organization were less likely than others to belong to an independent practice association and vice versa.
Managed Care Contracts and Hospitals
We found that radiology practices typically had little or no involvement in
the managed care negotiations of hospitals. The limited role of radiology
practices in their hospitals' managed care negotiations might be thought to
reflect the reality that hospitals predominantly negotiated only their own
payment—the technical fee. However, radiologists were only slightly more
involved when the hospital negotiated the global fee—that is, when the
radiologists' own income was at least indirectly at stake. Additional analysis
of the data suggests that the low level of radiology practice participation in
managed care contract negotiation by hospitals may be at least partly the
choice of many practices. Of the 58% of hospital-affiliated multiradiologist
practices with no involvement in contract negotiation, a full 78% indicated
that the practice either fully or partly determined whether the practice
participated in the contract negotiation. Thus, the small role of radiology
practices in hospital managed care negotiations reflects what the parties
prefer, not a restriction on participation forced by hospitals.
Strengths and Weaknesses of the Data
The survey data presented here, like all survey data, have strengths and
weaknesses. One strength is the large national sample of radiology practices
(n = 970) for the mailed survey, which constituted 23% of all
practices performing diagnostic radiology
[9]. The high usable response
rate (66%) to the survey enhances our confidence in the findings. Whereas
previous ACR surveys of radiology practices have been restricted to
multiradiologist groups, the data presented here also included
single-radiologist (solo) practices. The weighting of the data to reflect the
known distribution of practices by geographic region and practice size
increases the likelihood that the findings presented were representative of
radiology practices in the United States at the time of the survey. The broad
scope of questions asked in the survey enabled us to investigate a range of
practice characteristics potentially related to managed care. The survey of
radiology practices included two measures of managed care at the practice
level, enabling us to investigate both objective and subjective impact
(respectively, percentage of practice in managed care and perceived impact of
managed care). Inclusion of local area data external to the practice provided
a third, contextual measure of managed care: HMO penetration rate for the
local area.
Our data also have limitations. The original sampling frame for the survey was based on ACR members' responses to a prior mailing regarding regulatory compliance [9]. Radiology practices with no ACR members presumably were missing from the sample.
The individuals to whom the survey was mailed had been identified to the ACR as the heads of their practices. However, the existence of out-of-scope responses shows the list of addressees was imperfect. Also, the practices' business managers may have completed part or all of the questionnaire, particularly quantitative questions such as percentage of HMOs. The cover letter sent with the questionnaire suggested using the business manager extensively. To the extent that someone other than the practice head answered subjective questions, such as perceived impact of managed care, answers may not accurately represent the view of the practice head.
Answers regarding percentage of the practice that was Medicare, HMO, PPO, and so forth may be somewhat inaccurate because the respondent had imperfect knowledge and because some patients are in Medicare or Medicaid HMOs and it is not obvious how to classify them. We eliminated the worst inaccuracies by our requirement that the sum of the reported percentages had to be not too far from 100%. The small fraction (3%) of data that had to be eliminated indicates inaccuracies were not too severe. It is standard practice to ascertain percentage of managed care by self-report (see for example [2]). Validation would be possible only by on-site audit, a process so expensive that, to our knowledge, it is never used. Our validation based on the total of reported percentages is a quality control tool not applied by all researchers.
In regression analysis, the statistical effect of inaccuracies in independent variables is to attenuate relationships. Thus, to the extent the inaccuracies described in the previous paragraph are present, the true association of percentage of managed care with higher collection rates and of perceived impact of managed care with lower collection rates is stronger than our findings indicate.
Because a cross-sectional survey captures a snapshot of a single point in time, the causal nature of the relationships observed is not always clear. Therefore, we were cautious in attributing cause in the discussion of relationships we observed.
Managed care has changed since 1999 and therefore individual numeric values, such as the average percentage of managed care, almost certainly have changed some. However, relationships—for example, the high level of managed care involvement among multispecialty groups and the lack of association between percentage of managed care and various negative consequences—probably have not changed.
Conclusion
Data from the 1999 ACR Survey of Practices show that radiology practices
vary considerably in the extent of their involvement in managed care and that
many of the negative outcomes most feared by radiologists regarding the impact
of managed care on the business, professional, and organizational aspects of
practices had not materialized by the end of the 1990s. The data presented
here will provide a useful baseline for monitoring future trends, many of
which will be measured in a survey that will become available 1–2 years
from now.
Acknowledgments
We thank Douglas Wholey for construction of the data on local area HMO
penetration rates and for providing those data to us. Bruce Hauser, chair of
ACR's Commission on General and Pediatric Radiology, provided helpful comments
on an earlier draft.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. H. Sunshine, R. S. Lewis, and M. Bhargavan A Portrait of Interventional Radiologists in the United States Am. J. Roentgenol., November 1, 2005; 185(5): 1103 - 1112. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Hogan and J. H. Sunshine Financial Ratios in Diagnostic Radiology Practices: Variability and Trends Radiology, March 1, 2004; 230(3): 774 - 782. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
