Health Insurance and Medical Care: Physician Services under Managed Care

Distributed by Penny Hill Press

98-352 EPW

Health Insurance and Medical Care:

Physician Services under Managed Care

March 30, 1998

Patrick Purcell

Analyst in Social Legislation

Education and Public Welfare Division

ABSTRACT

Based on data from a 1995 national survey of physicians, this CRS Report compares

the frequency of diagnostic services and therapeutic services during office visits to physicians

who participated in health maintenance organizations (HMOs) and those who did not. The

report also describes the methods employed by managed care organizations (MCOs) to reduce

unnecessary use of health care services, and summarizes the findings of previous research on

the effect of managed care on health care utilization and outcomes. Results of the analysis

of physician office visits in 1995 are then displayed in a series of tables that show the

percei~tageof visits during which specific classes of services were performed by physicians

who participated in each of four kinds of health insurance arrangements. Statistical analysis

of the results indicates that while physician participation in an HMO had a significant positive

effect on the proportioil of visits during which these services were performed, the patient's

health status had the largest impact anlong all variables studied on the likelihood of receiving

a diagnostic or therapeutic service during an office visit.

Updates: No updates are planned for this report.

Health Insurance and Medical Care:

Physician Services under Managed Care

Summary

The financing and delivery of health care services in the U.S. have undergone

significant changes in recent years. In response to rapidly rising health care costs,

insurers, physicians, and employers have formed a variety of managed care

organizations (MCOs) in an effort to achieve greater efficiency Much has been

written about the potential savings in national health expenditures that may be realized

from the transition to managed care, but less is known about how this transition may

affect the specific medical services that patients receive. Studies that focus on the

pntienf 's insurance would be likely to find differences in treatments between patients

of comparable health status in prepaid and fee-for-service (FFS) plans only if

physicians differentiated their treatment decisions based on the kind of health

insurance by which the patient was covered. This Congressional Research Service

(CRS) Report adds to the literature on medical services under managed care by

analyzing data from a recent national survey of physicians. It compares the frequency

with which several classes of medical services were performed during physician office

visits classified according to the kind of insurance plans that the physician accepted.

This CRS Report makes several findings

a

In 1995, 84% of all office visits to primary care physicians by people under

age 65 with employer-group or other private health insurance were made to

physicians who participated in both managed and non-managed health

insurance plans

Diagnostic and screening services were performed more frequently during

visits both to pure FFS and pure HMO physician ofices than during visits to

physicians who accepted multiple kinds of insurance.

a

Both male and female patients were more likely to receive diagnostic or

screening services in a visit to an HMO-only physician ofice than during a visit

to a physician ofice that participated in a PPO, but the difference from the

overall average was greater for men than for women

a

Therapeutic and preventive services were performed more frequently during

ofice visits to HMO-only physicians than during visits to physicians who

accepted multiple kinds of insurance. For people who had no chronic medical

conditions, the likelihood of receiving therapeutic or preventive services did

not vary significantly according to the type of insurance accepted by the

physician.

Men were much more likely to receive therapeutic or preventive services in an

HMO-only office setting than in physician offices that accepted other kinds of

insurance. Women were equally likely to receive these services in HMO-only

or fee-for-service-only office settings.

Contents

Managing Health Care Use and Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.

The Financial Relationship with the Health Care Provider . . . . . . . . . . 2

Substituting Less Expensive Care for More Expensive Care . . . . . . . . 3

Reducing the Demand for Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.

Access to Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.

Health Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.

Data and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.

Results: Services Provided During Physician Office Visits . . . . . . . . . . . . . 8

Physician Ofice Visits: Patient Demographic Characteristics (Table 1)8

Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. .

Sex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. .

Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. .

Health Status: Chronic Medical Conditions . . . . . . . . . . . . . . . . . 9

Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9.

UrbanRural Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.

New or Previously Established Physician-Patient Relationship . . 9

Performance of Diagnostic or Screening Services (Table 2) . . . . . . . 11

Diagnostic and Screening Services: Differences by Sex (Table 3) . . . 11

Performance of Therapeutic or Preventive Services (Table 4) . . . . . . 12

Therapeutic or Preventive Services: Differences by Sex (Table 5) . . . 13

Multi-variate Analysis: Diagnostic and Screening Services . . . . . . . . 14

Age, Sex, and Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

.

Region and Urban/Rural Location . . . . . . . . . . . . . . . . . . . . . . 15

Health Status; New or Established Physician-Patient Relationship 15

Insurance Plans Accepted by Physician . . . . . . . . . . . . . . . . . . .15

Multi-variate Analysis: Therapeutic and Preventive Services . . . . . . . 16

Age, Sex, and Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

Region and Urban/Rural Location . . . . . . . . . . . . . . . . . . . . . .16

Health Status; New or Established Physician-Patient Relationship 17

Insurance Plans Accepted by the Physician . . . . . . . . . . . . . . . . 17

.

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18

Appendix I: Standard Errors of Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

.

Appendix 11: Medical Services and Medical Conditions . . . . . . . . . . . . . . . . . . 24

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

..

List of Tables

Table 1. Office Visits to Primary Care Physicians by Privately Insured People

UnderAge65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10

.

Table 2 . Diagnostic and Screening Services, Privately Insured Patients Under

Age65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

.

Table 3 . Diagnostic and Screening Services by Sex of Patient . . . . . . . . . . . . . 12

Table 4 . Therapeutic and Preventive Services, Privately Insured Patients Under

Age65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13.

Table 5. Performance of Therapeutic or Preventive Services

by Sex of Patient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

.

Table 6 . Diagnostic and Screening Services. Privately Insured Patients Under Age 65:

Logistic Regression Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Table 7 . Therapeutic and Preventive Services. Privately Insured Patients Under Age

65: Logistic Regression Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

.

Table A1 . Standard Errors for Estimates Displayed in Table 1 . . . . . . . . . . . . 21

Table A2 . Standard Errors for Estimates Displayed in Table 2 . . . . . . . . . . . . 22

Table A3 . Standard Errors for Estimates Displayed in Table 3 . . . . . . . . . . . . 22

Table A4 . Standard Errors for Estimates Displayed in Table 4 . . . . . . . . . . . . 23

Table A5 . Standard Errors for Estimates Displayed in Table 5 . . . . . . . . . . . . 23

Health Insurance and Medical Care:

Physician Services under Managed Care

Managed care comes in many forms, but all managed care organizations (MCOs)

share two basic goals. The most hndamental reasons for managing the delivery of

health care are to control costs while providing necessary and appropriate medico1

services. The number and variety of MCOs reflect the multitude of approaches to

achieving these objectives.' Although much has been written about the potential

savings in national health expenditures that may be realized from the transition to

managed care, less is known about how this transition may affect the specific medical

services that patients receive. If medical care services differ between managed care

and non-managed care or among the different kinds of MCOs, the reasons could be

related to the physicians' treatment practices, characteristics of the patients being

treated, or both. This information can be usehl both to users of health care services

and to policy-makers who must decide priorities in selecting and managing the

methods of financing and delivering health care for millions of beneficiaries in public

programs such as Medicare and Medicaid.

This CRS Report uses data &om a national survey of physicians to compare the

frequency with which several kinds of medical services were performed during office

visits in 1995 to physicians who participated in various kinds of MCOs and visits to

physicians who did not participate in managed care The results presented here are

based on physician office visits by people under age 65 who were covered by

employer-group or other private health insurance We have focused on privately

insured, non-elderly patients because it is among this segment of the population that

the transition to managed care has been occurring for the longest period of time, and

because the analysis is not complicated by issues of eligibility, enrollment, financing

and regulation that pertain specifically to the Medicare and Medicaid programs

Nevertheless, the experience of patients and physicians in the delivery of office-based

medical care under privately-financed MCOs may hold important lessons that can be

used to make the transition to managed care in the Medicare and Medicaid programs

more efficient and equitable for all interested parties

1

For a description of the ways in which health care providers and insurers have formed

MCOs see CRS Report 97-913, ManagedHealth Care: A Primer, by Jason S. Lee.

Some of these individuals also may have been covered by Medicare or Medicaid, but the

office visits included in this report all were covered by a private health insurance plan.

Managing Health Care Use and Costs

Arrangements for managing the financing and delivery of health care have

become more common in recent years, in part because the cost of health care rose

faster than the prices of most other goods and services from the 1960s through the

mid-1990s. Managed care seeks to make the provision of health care services more

efficient by using no more than the resources necessary to deliver care, while meeting

the community's prevailing standards of quality. The challenge faced by any MCO,

therefore, is one of controlling the cost of health care while maintaining the quality

of care. MCOs attempt to control costs by eliminating unnecessary or inappropriate

services. To do this they can:

create af2nancial relafionship with the health care provider that encourages

efficient use of resource^,^

substitute less expensiile care (for instance, office-based care) for more

expensive care (such as inpatient hospital care), and

reduce the demand ,for health care services by promoting good health

practices among those who are enrolled in the managed care plan.

The Financial Relationship with the Health Care Provider. The two most

common types of payment for medical care are "fee-for-service" (FFS) and

prepayment, which is also called "capitation" As its name implies, in a FFS

reimbursement arrangement the health care provider bills the patient for each

individual service provided. If the patient is insured, the insurer will usually pay some

or all of the bill, generally after the patient has paid the first few hundred dollars of

medical bills incurred during a calendar year, which is called a deductrble amount.

Insurance that pays expenses as they are incurred for each office visit, hospital stay,

or other episode of care is called lndemnrty msurance

Some indemnity insurers have negotiated special low reimbursement rates with

providers of health care. In exchange for these lower rates, the insurer designates the

health care professional as a "preferred provider." Patients who visit an insurer's

preferred providers are usually rewarded with a lower copayment (the share of the

total bill that is the patient's responsibility) than if they visit providers who are not

members of this network. Preferred provider organizations, or "PPOs," are

considered to be MCOs because they use financial incentives to induce health care

providers to deliver services efficiently. The financial incentives used by PPOs can

include amounts added to or withheld from the FFS reimbursement paid to the

provider. Inefficient providers may be dropped from the PPO, creating another

incentive for the provider to eliminate unnecessary services.

Under the FFS method of reimbursing health care providers, each additional test,

procedure, or office visit increases the provider's revenue, creating a financial

incentive for health care professionals to provide more services. This incentive exists

See also CRS Report 97-482, Managed Health Care: The Use of Financial Incentives by

Jason S. Lee and Beth C. Fuchs.

even in a PPO, although it may be tempered by the possibility that the provider could

be dropped from the PPO Another type of MCO, the health maintenance

organization, or HMO, reduces the financial incentive to provide additional care either

by employing the physician directly on a salary basis or by "prepaying" the physician

a fixed amount per month for each of the HMO's members in exchange for providing

all (or most) of the health care services that they require In the so-called "staffmodel" HMO, the physicians are employed directly, and their medical practice

patterns can by monitored and rewarded or sanctioned, just like the performance of

an employee in any other kind of firm. In "group-model" HMOs and in another kind

of HMO called an "independent practice association" (IPA), the HMO contracts with

physician groups or independent physician practices to provide health care services

to the HMO's members, usually in exchange for a monthly "capitation payment "4 If

the total cost of care provided to the members of the HMO is less than the sum of

these capitation payments, the physician group keeps the difference as profit If costs

exceed the sum of the capitation payments, the physicians suffer a financial 1 0 ~ s . ~

Substituting Less Expensive Care for More Expensive Care. MCOs also are

able to reduce costs by employing primary care physicians as "gatekeepers" t o

regulate patients' access to more expensive specialists, and by providing as much care

as possible in physician offices, clinics, and hospital outpatient departments rather than

in an inpatient hospital setting. Lower rates of hospitalization and shorter stays in

hospitals associated with MCOs - specifically with HMOs - have been shown t o

be an important source of savings in MCOs compared with traditional indemnity

insurers

Reducing the Demand for Care. By promoting good health practices that

emphasize early intervention and preventive medicine, such as immunizations,

screening for specific conditions, and regular check-ups, MCOs may lower the total

cost of health care by catching incipient illnesses when they can be treated with the

greatest effectiveness and at the least expense In such cases, more frequent use of

health care services in the near-term may result in reduced need for medical services

in the long-term

Even HMOs that pay most primary-care physicians on a capitation basis will often

reimburse specialists FFS. Members of an HMO may sonletimes see non-participating

physicians who are reimbursed on a FFS basis if the HMO offers a "point-of-service" (POS)

option.

5

In 1996, 74% of HMO contracts with primary-care physicians provided for prepaid,

capitated reimbursement, according to Hoechst-Marion-Roussel's HMO-PPO Digest

To date, reductions in health care costs resulting from managed care have been attributed

mainly to reductions in the number and length of inpatient hospital stays. See, for example,

KPMG Peat Marwick. The Impact ofManaged Care on U S . Markets. Washington DC,

1996 and; Richardson, James C. Decline in Hospital Utilization and Cost Inflation Under

Managed Care in California. Journal of the American Medical Associat~on,v. 276, no. 13,

October 2, 1996.

7

Early diagnosis and treatnlent of illnesses may be desirable for their public health benefits,

but they may not result in savings to national health expenditures. A person who detects a

medical condition at age 70, for example, and has it treated successfully, possibly preventing

(continued...)

Previous Research on Ambulatory Services under Managed Care

Access to Care. Numerous studies published in recent years have examined

issues of access to care, intensity of service utilization, and quality of care in managed

and non-managed health insurance plans A review of the literature evaluating

managed care plan performance between 1980 and 1993 found that, compared to

indemnity insurance plans, HMOs had "the same or more physician office visits per

enrollee, less use of expensive procedures and tests, greater use of preventive

services, [and] mixed results on outcomes A more recent literature review by the

same authors found that "there was no overall pattern to physician use or expenditure

results" in 10 studies published between 1993 and 1997 Another recent study found

that as the average number of inpatient hospital days among patients in HMOs has

been declining, the frequency of physician office visits has been rising 'O Data from

the 1994 Health Interview Survey analyzed by CRS show that the frequency at which

people in HMOs visited physician offices did not differ substantively from the rate for

people in traditional indemnity insurance plans These data show that 15 0% of

privately insured people under age 65 in HMOs had made at least one office visit in

any given 2-week period in 1994, compared to 14 1% of people covered by indemnity

insurance plans that paid providers on a FFS basis "

"'

One reason that the frequency of physician visits differs only slightly between

people enrolled in HMOs and those with FFS insurance is that average health status,

as measured by the percentage of people who have chronic health problems differs

very little between these two groups. In the 1994 National Health Interview Survey,

37% of people enrolled in indemnity insurance plans and 37% of those in HMOs

reported that they had a medical condition of any kind. Among those covered by

indemnity plans, 30.7% reported one or more chronic medical conditions, versus

29.9% of those in HMOs." The frequency of office visits was similar between the

' (...continued)

premature death, may eventually suffer another illness or series of illnesses requiring medical

treatment and additional expenditures.

'Miller, R.H. and H.S. Luft "Managed Care Plan Performance Since 1980. Journal ofthe

American Medical Association, v. 271, no. 19, May 18, 1994.

Miller, R.H. and H.S. Lufi. "Does Managed Care Lead to Better or Worse Quality of

Care?" Health ASfairs, v. 16, no. 5, SeptemberIOctober 1997.

l o Wholey, D., et al. HMO Market Structure and Performance: 1985-1995.Health ASfairs,

v. 16, no. 6, NovemberlDecember 1997.

" Although small, this difference is statistically significant at the 95% confidence level, based

on a two-tailed t-test.

l 2 The extent, if any, to which HMOs experience favorable selection in terms of the health

status of their enrollees remains an open question. Data from the 1994 National Health

Interview Survey indicate little difference in the health status of people enrolled in' HMOs

compared to those in indemnity insurance plans. Taylor and colleagues (1995), reporting on

an analysis of data from the 1987 National Medical Expenditure Survey, also found "very

little evidence that individuals enrolled in HMOs were healthier than those in FFS plans."

Likewise, Fama and colleagues (1995) found little observable difference in the health status

(continued...)

HMO and FFS enrollees even among those who reported having a medical

condition.I3 Of the people who reported having an acute or chronic medical

condition, 34.8% of HMO enrollees had seen a doctor in a given 2-week period,

compared with 33.2% of those enrolled in indemnity insurance plans.14

Health Outcomes. Studies of outcomes between managed and non-managed

health care plans also have found mixed results Clement and others (1994) found

that Medicare beneficiaries in HMOs who reported either joint pain or chest pain were

less likely to have seen a specialist for care, to have a follow-up recommended, or to

have their progress monitored Less improvement in symptoms was found among the

HMO enrollees in one of four outcomes studied In a 4-year study of chronically ill

patients, Ware and others (1996) found that physical and mental health outcomes did

not differ between FFS and HMO plans for the average patient, but they did for

certain sub-groups of patients Among elderly patients, declines in physical health

were more common among those in HMOs than in FFS plans Outcomes tended to

be better in FFS plans among poor patients, but better outcomes were found in HMOs

for non-poor patients Yelin, Criswell, and Feigenbaum (1996) examined data on

health outcomes over periods of up to 11 years for patients with rheumatoid arthritis

in FFS insurance plans and prepaid group practices They found that the "two groups

did not differ on any outcome measure on either an annual or long-term basis "

Data and Methods

Access to ambulatory medical care services oRen is measured in terms of the

frequency of contacts with a physician or other health care professional Another

measure of access is the likelihood that a physician office visit will result in a specific

service being provided If the likelihood that a specific service will be performed

during an office visit appears to differ according to the kinds of insurance accepted

by the physician, we would like to know the extent to which this variation reflects

differences in the patients being treated, the physicians' practice patterns (including

the influence of managed care), or a combination of these factors This CRS Report

uses data from a nationally representative survey of physicians t o examine the

frequency with which particular classes of medical services were performed during

office visits to physicians who participated in different kinds of insurance

arrangements The report focuses on two broad classifications of services identified

(...continued)

of people in HMOs and those in other insurance plans. Helllllger (1995), however, in a review

of the literature on the subject reported that a majority of studies found that HMOs and other

plans with closed provider networks experienced favorable selection bias among their enrollees

when compared to people enrolled in non-HMO plans.

l 3 Among people who reported no nledical conditions, 3.6% of those who were enrolled in an

HMO saw a physician in a 2-week period, compared with 3.0% of those covered by indemnity

insurance, not a statistically significant difference.

Although the differences in chronic conditions and 2-week ofice visits are not large enough

to suggest dramatic behavioral differences either by patients or physicians in HMOs compared

to indemnity plans, both are statistically significant at the 95% confidence level based on a

two-tailed t-test.

l4

in the National Ambulatory Medical Care Survey: diagnostzc and screening services

and therapeutic andpreventive services.

For this report, we analyzed the frequency with which specific services were

performed during physician office visits according to the kind of health insurance

plans accepted by the physician. Most prior studies have looked at the kind of

insurance that covered the individual patient; however, in 1995 more than four-fifths

of all office visits to primary care physicians by privately insured individuals under age

65 were made to physicians who saw some patients in managed health plans and some

with non-managed insurance. Studies that focus on the kind of insurance that covered

the patient would be likely to find differences in treatments or outcomes between

patients in prepaid and FFS plans only if a significant proportion of physicians

differentiated their treatment decisions based on the kind of health insurance by which

the patient was covered. This hypothesis is not generally supported by empirical

findings.'' The methodology employed for this analysis permitted us to see whether

the frequency of services differed between physicians who accepted only prepaid

insurance such as HMOs or only FFS insurance. This classification scheme also

allowed us to see whether the frequency with which certain categories of medical

services were performed differed between physicians in pure-HMO or pure-FFS

practices and physicians who see patients in both prepaid and FFS plans.

The analysis presented here focuses on ofice visits to physicians who are

primary-care providers because it is these physician offices in which managed care

principles have been most widely adopted. Likewise, in some of the tables that

follow, the sample has been restricted to ofice visits made by patients who were

reported by the physician to have no chronic medical conditions. This restriction can

help to make clear whether a difference in the frequency with which specific services

were provided was the result of differences in physician practice patterns between

managed care and non-managed care; differences in the proportion of people enrolled

in each kind of plan who were in poor health; or a combination of these two factors.

Data on health care services provided during visits to ofice-based physicians

were collected as part of the 1995 National Ambulatory Medical Care Survey

(NAMCS) conducted by the National Center for Health Statistics, a division of the

U.S. Department of Health and Human Services. All non-federally employed officebased physicians who were primarily engaged in patient-care activities were eligible

to be included in the Ambulatory Medical Care Survey with the exception of

physicians whose specialties were anesthesiology, pathology, or radiology. Contacts

with patients by telephone, outside the physician's office, in hospital settings (inpatient

or outpatient), in institutional settings such as nursing homes, or office visits for

administrative purposes (such as paying a bill, filling out insurance forms, or dropping

'' For example, many physicians are unaware even whether a patient is covered by private

insurance or by a publicly financed program such as Medicaid. Research into state

compliance with the early, periodic, screening, diagnosis and testing (EPSDT) requirements

for children covered by Medicaid indicates that most physicians performing a physical exam

or well-child checkup are unaware of the child's source of health insurance coverage. (Elicia

Herz, Ph.D., MEDSTAT Group, Washington DC, personal communication with author.)

off a specimen) were not included in the survey. Of the 2,587 physicians contacted

in 1995, 73% participated in the survey.

The 1995 survey data consist of 36,875 patient record forms collected from

1,883 physicians who participated in the NAMCS. Each record represents one visit

by a patient to an office-based physician, and each has been assigned a patient visit

weight, which when aggregated across all visits, represent the estimated total of 697

million physician office visits in the U.S. in 1995. For this report CRS limited its

analysis to physician visits by people under age 65 with employer-group or other

private health insurance. This sub-sample comprised 27,067 patient record forms

(73.4% of the total), and represented 254.5 million physician ofice visits.

Of the 254.5 million office visits by people under age 65 with employer-group

or other private health insurance, 168.3 million visits (66.1%) were made to primarycare physicians, including general and family practitioners, internists, pediatricians,

and obstetricianslgynecologists. Just 10% of these visits were made to physicians

who accepted only insurance that reimbursed the physician on a FFS basis and who

did not participate in either a PPO or HMO. Only 6% ofthese ofice visits were made

to physicians who only saw patients who were enrolled in an HMO. Thus, relatively

small proportions of physician office visits in 1995 were made to physicians who had

no patients in managed care or to physicians who saw only patients who were enrolled

in the most strictly managed form of managed care. The vast majority of physician

office visits in 1995 - 84% of the total - were made to physicians who treated

. some patients who were enrolled in a managed care plan, and some who were not

Patient records on the NAMCS include both the expected source of insurance

and the expected source of payment The categories for expected source of insurance

are Blue CrossiBlue Shield, other private insurance, Medicare, Medicaid, worker's

compensation, other insurance, and unknown There are eight source-of-payment

categories on the survey instrument

blank with no source of insurance reported,

preferred provider option,

insured FFS,

HMO or other prepaid,

a

self-pay,

a

no charge,

e

other source of payment, and

type of payment was not specified but source of insurance was reported.

In this report, physician office visits have been grouped according to the sources

of payment reported on patient records during the week that each physician

participated in the NAMCS. We used the unique physician identifiers to classify

physicians into four groups, based on the kinds of insurance that covered patients seen

by the physician. The four groups of physicians were:

a

those who saw only patients whose insurance reimbursed the physician on a

FFS basis with no PPO discounts,

those who saw patients in both regular FFS plans and in PPO plans, but who

saw no patients who were enrolled in an HMO,

a those who saw patients enrolled in HMOs, in PPOs, and in non-PPO plans,

a

those who saw only patients who were enrolled in an HM0.16

This classification allows us to discern not only whether services differed

between managed care and non-managed care, but also whether the frequency of

services provided during visits to "mixed insurance" physician offices was more like

the frequency seen in visits to HMO-only physician offices or like that which occurred

during visits to physicians who accepted only pure FFS payments.

Results: Services Provided During Physician Office Visits

Physician Office Visits: Patient Demographic Characteristics (Table 1).

Patient demographic and health characteristics, geographic variations in

physician practice patterns, and whether the physician has seen the patient before all

may influence diagnostic and treatment procedures. These characteristics are arrayed

in Table 1 according to the proportion of office visits made by people in each

category to physicians who accepted each of four kinds or combinations of health

insurance plans.

Age. Children and young adults were less likely than adults 35 or older to have

made an office visit to a physician who accepted only FFS insurance arrangements,

and were more likely than people in other age groups to have visited a physician office

that accepted both HMO and PPO insurance plans. People between the ages of 55

and 64 made up 18.1% of visits to physicians who accepted only FFS insurance

arrangements, significantly higher than the overall average of 10.0%; however, they

also comprised a higher-than-average share of visits to physicians who treated only

patients enrolled in HMOs.

Sex. In 1995, 107 million out of 168 million office visits (63.5%) to primary

care physicians by privately insured people under age 65 were made by female

patients. Male and female patients had very similar probabilities of visiting a physician

The data collected on the Anlbulatory Medical Care Survey do not differentiate between

grouplstaff model HMOs and IPA-model HMOs, a distinction that could increase the

explanatory power of these data with respect to differences among HMOs in the frequency

with which specific services were provided during office visits. Nevertheless, because neither

kind of HMO typically uses FFS reimbursement, the NAMCS data allow us to distinguish

between kinds of insurance payment that respond to increases in the number of patient visits

or volume of services and those that do not.

l6

who treated only HMO patients, but males made up an above-average share of all

visits to physicians who accepted only FFS insurance plans.

Race. Both white and non-white patients had similar likelihoods of visiting a

primary care physician who saw only HMO enrollees, but non-white patients had an

above-average probability of visiting a physician who accepted only FFS insurance

reimbursement.

Health Status: Chronic Medical Conditions. Office visits to physicians who

accepted only FFS reimbursement included a higher-than-average proportion of visits

made by patients with chronic medical conditions, while visits to physician offices that

accepted both FFS and PPO plans included a lower-than-average proportion of visits

made by chronically ill patients." Visits to FFS-only physician offices made up 10%

of all visits by privately insured patients under age 65, but they comprised 14.8% of

all visits made by patients with one or more chronic medical conditions. Visits to

physician offices that participated in both FFS and PPO plans made up 16.1% of all

visits by privately insured patients under age 65, but they comprised 14.2% of all

visits made by patients with one or more chronic medical conditions.

Region. There were significant regional differences in the proportion of office

visits that were made to physicians who participated in an HMO In the West, 11 1%

of all visits were made to physicians who treated only patients enrolled in an HMO,

while in no other region were more than 4 2% of visits made to HMO-only physician

offices In both the Northeast and the West, more than 80% of all office visits were

made to physicians who saw some HMO patients, usually along with PPO patients

In the Midwest and South, fewer than 70% of all office visits were made to primary

care physicians who participated in at least one HMO

Urban/Rural Location. In 1995, 78.7% of office visits in urban areas were

made to physicians who treated at least some patients who were enrolled in an HMO.

Only 7.1% of office visits to physicians in urban areas were to offices that accepted

only FFS reimbursement. In contrast, just 48.1% of office visits in rural areas were

made to physicians who treated at least some patients who were in an HMO. More

than a quarter of all office visits in rural areas were made to physicians who accepted

only insurance plans that reimbursed the provider on a FFS basis.

New or Previously Established Physician-Patient Relationship. Among office

visits to primary care physicians by privately insured people under age 65, 90.2%

were made to a physician who had seen the patient before (15 1,818 out of 168,289).

Office visits by patients that the physician had never seen before were most likely to

have occurred in among physicians who participated in both HMO and PPO insurance

plans. While two-thirds of all office visits by privately insured people under age 65

were made to physicians who saw patients in both HMO and PPO plans, visits to

l7

In the National Ambulatory Care Survey, chronic conditions are defined to include arthritis,

arteriosclerosis, chronic obstructive pullnonary disorder, renal disease, clinical depression,

diabetes, human immunodeficiency vims/aquired immune deficiency syndrome (HIVIAIDS),

hyperactivity, hypertension, and obesity.

HMO and PPO offices comprised three-fourths of all visits in which the physician had

not seen the patient before.

Table 1. Office Visits to Primary Care Physicians by Privately Insured

People Under Age 65

Percentage of visits that were to

phjisicians who accepted:

Visits

FFS and H M O and H M O

(000s) FFS only PPO

PPO

only

Total

Patient demographics:

Under age 18

Ages 18 to 34

Ages 35 to 54

Ages 55 to 64

59,144

36,224

53,413

19,508

Z9%

8.4%

10.5%

18.1%

18.4%

11.8%

16.4%

16.3%

71.5%

72.6%

65.6%

56.8%

2.3%

73%

Z4%

8.8%

100%

100%

100%

100%

Female

Male

106,942

61,347

8.8%

12.1%

15.7%

16.9%

69.8%

65.4%

5.8%

5.7%

100%

100%

White

Black or other

147,423

20.865

9.5%

13.8%

16.9%

10.2%

67.9%

70.1%

5.7%

60%

100%

100%

Has Chronic Condition(s) 36,434

No Chronic Conditions

131,853

14.8%

8.7%

14.2%

16.6%

64.0%

69.3%

7.0%

5.4%

100%

100%

Physician Locatron:

Northeast

Midwest

South

West

29,070

40,991

52,684

45,544

10.9%

10.3%

9.3%

101%

4.5%

19.9%

25.6%

9.1%

81.5%

65.6%

61.4%

69.8%

3.2%

4.2%

3.8%

11.1%

100%

100%

100%

100%

Patient Seen Previously

Not Seen Previously

151,818

16,470

10.3%

7.5%

16.7%

10.7%

67.4%

75.2%

5.6%

6.6%

100%

100%

Total

168.289

10.0%

16.1%

68.2%

5.7%

100%

Urban

Rural

Source: CRS analysis of data fro111the 1995 NAMCS

Note: Percentages that are significantly different from the column average at 95% or greater

confidence are shown in boldface type.

Performance of Diagnostic or Screening Services (Table 2). Diagnostic or

screening services were performed during 74.3% of all ofice visits to primary care

physicians in 1995, but the frequency varied according to the type of insurance

accepted by the provider. (Diagnostic and screening services are defined in Appendix

11). Visits to HMO-only physicians had the highest probability of a diagnostic or

screening service, at 85% of visits. Ofice visits to FFS-only physicians had the

second highest probability of a diagnostic or screening service (78%). Oflice visits

to physician practices that accepted both FFS and PPO plans had lower-than-average

rates of screening and diagnostic services. The pattern was similar among patients

with no chronic conditions, except that among this group the proportion visits to FFSonly physicians during which diagnostic and screening services were performed was

not statistically different from the average for all ofice visits by patients with a

chronic condition.

Table 2. Diagnostic and Screening Services, Privately Insured Patients

Under A g e 65

Were diagno.~tic/screenrngservices performed?

Type of insurance

accepted by physician

Visits (000s)

Yes

No

Total

FFS only

FFS or PPO

HMO or PPO

HMO only

16,848

27,075

114,703

9,662

77.9%

70.5%

73 8%

84.8%

22.1 %

29.5%

26 2%

15.2%

100%

100%

100%

100%

Total

168,288

74.3%

25.7%

100%

Privately Insured Patients Under Age 65 Who Had No Chronic Medical Conditions

Were diagnostic/screening services performed?

FFS only

FFS or PPO

HMO or PPO

HMO only

Total

Visits (000s)

11,459

21,915

91,365

7,115

Yes

No

Total

71.3%

65.3%

69.8%

81.4%

28.7%

34.7%

30.2%

18.6%

100%

100%

100%

100%

131,854

69.8%

30.2%

100%

Source: CRS analysis of data from the 1995 NAMCS

Note: Percentages that are sigllificantly different from the colutnn average at 95% or greater

confidence are shown in boldface type.

Diagnostic and Screening Services: Differences by Sex (Table 3). Overall,

diagnostic or screening services were more likely to be performed during ofice visits

made by female rather than male patients, (78% vs. 68%) Both female and male

patients were most likely to have received diagnostic or screening services during a

visit to an HMO-only physician, but the difference was about twice as large for males

(83% in pure-HMO offices vs. 67% in other offices ) as for females (86% in pureHMO offices ~ ~ 7 in

7 other

% offices). Male and female patients were equally likely

to receive diagnostic or screening services in physician offices that accepted only

traditional FFS insurance (78% for both sexes). Among female patients, the

likelihood of receiving these services was significantly different from the overall

average only for visits to pure-HMO physician offices. Among male patients, the

likelihood of having a diagnostic or screening service performed was significantly

higher than average in both pure-HMO and pure-FFS physician offices, and it was

significantly below average in offices that accepted both FFS and PPO plans.

Table 3. Diagnostic a n d Screening Services by Sex of Patient

Female patients

Were diagnostic/screening services performed?

Type of insurance

accepted by physician

FFS only

FFS and PPO

HMO and PPO

HMO only

Total

Total visits

(000s)

9,408

16,738

74,602

6,194

Yes

No

Total

77.6%

75.7%

77.8%

85.6%

22.4%

24.3%

22.2%

14.4%

100%

100%

100%

100%

106.942

77.9%

22.1 %

100%

Male patients

Were diagnostic/screening sewices pevformed?

Type of insurance

accepted by physician

FFS only

FFS and PPO

HMO and PPO

HMO only

Total

Total visits

(000s)

7,440

10,337

40,101

3,469

Yes

No

Total

78.3%

62.0%

66.3%

83.3%

21.7%

38.0%

33.7%

16.7%

100%

100%

100%

100%

61,347

68.0%

32.0%

100%

Source: CRS analysis of data from the 1995 NAMCS.

Note: Percentages that are significantly different from the column average at 95% or greater

confidence are shown in boldface type.

Performance of Therapeutic or Preventive Services (Table 4). Only 36.5%

of office visits to primary care physicians in 1995 included performance of at least one

therapeutic or preventive service. (Therapeutic and preventive services are defined

in Appendix 11.) As was also the case with diagnostic and screening services, office

visits to physicians who treated only patients enrolled in an HMO had the highest

frequency of therapeutic and preventive services (43.4%), and visits to physicians who

accepted both FFS and PPO insurance plans included performance of these services

at a rate that was significantly below the overall average.

One possible explanation for differences in the proportion of visits during which

therapeutic or preventive services were administered would be differences in the

health status of patients seen by physicians in each of the four insurance categories.

The bottom panel of Table 4 shows office visits by people who were reported by the

physician to have no chronic medical conditions. Among patients with no chronic

medical conditions, the rate of therapeutic and preventive services did not differ

significantly from the overall average for any of the four insurance classifications.

Table 4. Therapeutic a n d Preventive Services, Privately Insured

Patients U n d e r Age 65

Were therapeutic or preven five services performed?

Type of insurance

accepted by physician

FFS only

FFS or PPO

HMO or PPO

HMO only

Total

Yes

No

Total

16,848

27,075

114,703

9,662

38.6%

32.2%

36.7%

43.4%

61.4%

67.8%

63.3%

56.6%

100%

100%

168,288

36.5%

63.5%

100%

Visits (000s)

100%

100%

Privately Insured Patients Under Age 65 Who Had No Chronic Medical Conditions

Were therapeutic or preventive services performed?

Type of insurance

accepted by physician

FFS only

FFS or PPO

HMO or PPO

HMO only

Total

Visits (000s)

Yes

No

Total

11,459

21,915

91,365

7,115

30.2%

28.7%

32.2%

35.7%

69.8%

71.3%

67.8%

64.3%

100%

100%

100%

100%

131,854

31.6%

68.4%

100%

Source: CRS analysis of data from the 1995 National Ambulatory Medical Care Survey.

Note: Percentages that are significantly different from the column average at 95% or greater

confidence are shown in boldface type.

Therapeutic or Preventive Services: Differences by Sex (Table 5). There

was little difference in the proportion of office visits by male and female patients

during which therapeutic or preventive services were performed. These services were

performed during 37.5% of physician office visits by men and 36% of visits by

women. Among female patients, therapeutic or preventive services were equally

likely to be administered during a visit to an HMO-only or FFS-only physician

practice (41%), while visits to physician offices that participated in both FFS and PPO

plans included therapeutic and preventive services at a rate significantly lower than

the overall average - (30.7% versus 36.0%). Among men, visits to HMO-only

physicians had a higher-than-average frequency of therapeutic and preventive services

(48%), while visits to all other physician offices included these services at rates near

the overall average of 37.5%.

Table 5. Performance of Therapeutic or Preventive Services

by Sex of Patient

Female patients

Were therapeutic or preventive services performed?

Type of insurance

accepted by physician

Visits (000s)

Yes

No

Total

FFS only

FFS and PPO

HMO and PPO

HMO only

9,408

16,738

74,602

6,194

40.9%

30.7%

36.2%

40.8%

59.1 %

69.3%

63.8%

59.2%

100%

100%

100%

100%

Total

106,942

36.0%

64.0%

100%

Male patients

Were therapeutic or preventive services performed?

Type of insurance

accepted by physician

Visits (000s)

Yes

No

Total

FFS only

FFS and PPO

HMO and PPO

HMO only

7,440

10,337

40,101

3,469

35.7%

34.5%

37.7%

47.9%

64.3%

65.5%

62.3%

52.1%

100%

100%

100%

100%

Total

61,347

37.5%

62.5%

100%

Source: CRS analysis of data from the 1995 National Ambulatory Medical Care Survey.

Note: Percentages that are significantly differentfrom the column average at 95% or greater

confidence are shown in boldface type.

Multi-variate Analysis: Diagnostic and Screening Services. To evaluate the

relationship of each the variables described in Table 1 through Table 5 to the

probability that a physician office visit included a diagnosticiscreening or

therapeuticipreventive service, we processed the data though a regression model18.

This form of analysis shows the relative impact of each specified characteristic of the

patient or physician while all of the other factors are held constant.

Age, Sex, and Race. Other things being equal, the probability that a privately

insured individual under age 65 had a diagnostic or screening service performed

during a physician office visit increases by about 4.6% for each additional year of age,

and a female patient was about 33% more likely to have a diagnostic or screening

service performed during an office visit than was a male patient. (See the column

labeled "Odds Ratio.") A patient's race (defined here as white or nonwhite) was

unrelated to the likelihood that he or she had a diagnostic or screening service

performed, once other patient and physician characteristics were taken into account.

Region and UrbadRural Location. Both the region and urban or rural location

of a physician office had a statistically significant relationship to the probability that

a diagnostic or screening service was performed during a physician office visit.

Relative to physician office visits in the Midwest, office visits in the Northeast, South,

and West all had a greater likelihood of including a diagnostic or screening service,

with the probability ranging from 33% higher in the South to 47% higher in the

Northeast. Likewise, visits to physician offices in urban areas were about one-third

more likely than those in rural areas to include a diagnostic or screening service, other

factors being equal.

Health Status; New or Established Physician-Patient Relationship. Patients

who had one or more chronic conditions according to the patient records maintained

by the physician were much more likely to have undergone a diagnostic or screening

service during an office visit than patients who had no chronic conditions The rate

at which these services were performed was almost 80% higher for a patient with a

chronic condition when other characteristics were held constant Having seen a

patient during a previous visit was associated with a lower probability of receiving

diagnostic or screening services Patients who had been seen by the health care

provider previously were about 26% less likely to have had a diagnostic or screening

service performed than those who were being seen for the first time Office visits also

were more likely to include diagnostic and screening services if the provider was a

physician rather than a nurse practitioner, registered nurse, or other provider,

however, visits to these other providers accounted for only 4% of all office visits

Insurance Plans Accepted by Physician. For the regression analysis, we used

office visits to physicians who accepted only non-PPO FFS insurance plans as the

reference group for estimating the probability of a diagnostic or screening service

being performed during an office visit. Compared to the reference group, there was

no significant difference in the likelihood of a diagnostic or screening service being

performed during visits to physicians who accepted both FFS and PPO insurance

plans. Visits to HMO-only and HMOPPO physician offices, however, were both

associated with higher rates of diagnostic and screening services, than office visits to

pure-FFS offices, other things being equal. An office visit to a physician who

'*The results presented in this section are from a logistic regression model. We also

processed the data through a probit model and found substantively the same results.

participated in both PPO and HMO plans was about 20% more likely to include a

diagnostic or screening service than a visit to a FFS-only office, while the likelihood

of having these services performed during a visit to an HMO-only physician office

was about 40% higher than in a pure-FFS office, other things being equal.

Table 6. Diagnostic a n d Screening Services, Privately Insured Patients

U n d e r Age 65: Logistic Regression Results

Response Variable: Were diagnostic or screening services performed?

Weighted

Parameter

Standard

Analysis Variable

Mean

Estimate

Error Odds Ratio

-2.3615 ***

0.2572

Intercept

0.00202

1.046

29.1

0.0449 ***

Age

0.0670

1.329

0.6345

0.2843 ***

Sex (1 = female)

0.8778

-0.0551

0.1060

0.946

Race (1 = white)

0.1736

0.3835 ***

0.1043

1.467

Northeast

0.3241

0.2874 ***

0.0866

1.333

South

0.0931

1.386

0.2640

0.3262 ***

West

0.0916

1.333

0.8336

0.2876 ***

Urban Area

0.1099

1.793

0.2189

0.5840 ***

Has Chronic Condition(s)

-0.3066 ***

0.1146

0.736

Patient Seen Previously

0.9035

0.1499

6.138

0.9583

1.8145 ***

Provider is an M,D.

0.1265

1.174

0.1569

0.1600

Reg. FFS or PPO accepted

0.1083

1.201

0.6645

0.1828 *

HMO or PPO accepted

0.1921

1.397

0.0560

0.3341 *

Only HMO accepted

Source: CRS analysis of data from the 1995 NAMCS.

n = 6,028 observations

* significant at >= .10

*** significant at >= .O1

Multi-variate Analysis: Therapeutic and Preventive Services.

Age, Sex, and Race. When the response variable of the regression model was

changed to indicate whether therapeutic or preventive services were performed during

an office visit, the relationship of the patient's age was significant, but negative. Each

additional year of age reduced by about 1% the likelihood of a therapeutic or

preventive service being performed during an office visit. The signs of the variables

indicating the patient's sex (female = 1) and race ( white = 1) also were negative, but

neither of these variables was statistically significant.

Region and Urban/Rural Location. Relative to being located in the Midwest,

an office visit in the Northeast was more likely to include a therapeutic or preventive

service, and an office visit in the South was less likely to include these services.

Physician ofice visits in the West had effectively the same likelihood of including

therapeutic and preventive services as office visits in the Midwest. As was the case

CRS- 17

with diagnostic and screening services, therapeutic and preventive services were more

likely to be administered during a physician office visit in an urban area than in a rural

area, with the probability being about 23% greater in urban settings.

Health Status; Nerv or Established Physician-Patient Relationship. The

presence of one or more chronic medical conditions had a positive and statistically

significant relationship to the probability that therapeutic and preventive services were

performed during a physician office visit. Other things being equal, patients with

chronic conditions were almost three times as likely to have these kinds of services

performed as were patients with no chronic conditions. Office visits to a health care

provider who had seen the patient previously were less likely to include therapeutic

and preventive services than visits to a provider who had never before seen the

patient. An office visit with an M.D. was much more likely to include therapeutic and

preventive services than a visit with a primary-care provider who was not a physician.

Insurance Plans Accepted by the Physician. Relative to visits to physicians

who accepted only regular fee-for-service (FFS) insurance plans, neither visits to

providers that accepted both FFS and PPO or both HMO and PPO plans had a

significantly different likelihood of including a therapeutic or preventive service, other

things being equal As was also the case with diagnostic and screening services,

however, office visits to primary-care providers who accepted only HMO insurance

plans were significantly more likely than visits to a pure-FFS provider to include a

therapeutic or preventive service, once other factors were taken into account

Relative to an office visit to a provider who accepted only FFS insurance plans, a visit

to an office that accepted only HMO plans was about 37% more likely to include

delive~yof therapeutic or preventive services

Table 7. Therapeutic and Preventive Services, Privately Insured

Patients Under Age 65: Logistic Regression Results

Response Variable: Were therapeutic or preventive services performed?

Weighted

Parameter

Standard

Analysis Variable

Mean

Estimate

Error Odds Ratio

Intercept

Age

Sex (1 = female)

Race (1 = white)

Northeast

South

West

Urban Area

Has Chronic Condition(s)

Patient Seen Previously

Provider is an M.D.

Reg. FFS or PPO accepted

HMO or PPO accepted

Only HMO accepted

Source: CRS analysis of data from the 1995 NAMCS

n = 6,028 observations

* significant at >= .10

*** significant at >= .05

*** significant at >= .O1

Discussion. The data reported in Table 1 show that patients with particular

demographic and health characteristics are more likely t o have visited the offices of

primary-care physicians that accept certain kinds of insurance and less likely to have

visited physician office that accept other kinds of insurance. Pure FFS physician

offices had a smaller-than-average percentage of visits by children and young adults

and a greater than average percentage of visits by adults aged 55 to 64. Pure HMO

physician practices also had a lower-than-average proportion of visits by children, but

a greater-than-average proportion of visits but adults aged 35 to 64. Male patients,

non-white patients, patients with one or more chronic medical conditions, and patients

in rural areas each made up a greater-than-average proportion of visits to pure-FFS

physician offices. Visits to pure FFS offices had a lower-than-average percentage of

visits by patients whom the physician had never seen before, while offices that

accepted both HMO and PPO plans had a disproportionately large share of visits by

patients who were seeing a particular provider for the first time.

The data presented in Table 2 and Table 3 show consistently higher rates of

diagnostic and screening services being performed in pure-HMO physician offices

and that the relative difference from the average was much greater for men than for

women. In contrast, while men also were more likely to have received a diagnostic

or screening service in a visit to a FFS-only office and were less likely to have

received these services in a mixed FFSIPPO office, no such differences were found

for women. Overall, women were more likely than men to have received diagnostic

or screening services during a typical office visit, but among all visits by women the

likelihood of receiving diagnostic or screening services was significantly above

average only during visits to pure-HMO physician offices. Patients with no chronic

medical conditions were less likely to receive diagnostic or screening services than

those who had a chronic condition. Patients without a chronic condition were most

likely to receive diagnostic or screening services in an HMO-only office.

Table 4 and Table 5 show a different pattern with respect to therapeutic and

preventive services than Tables 1 and 2 showed for diagnostic and screening

services. Overall, the likelihood of receiving therapeutic or preventive services was

similar for male patients and female patients. Visits to pure-HMO offices had a

greater likelihood of including a therapeutic or screening services, while visits to

mixed FFS and PPO offices had a lower-than average likelihood of including these

services. The higher rate in HMO offices was statistically significant only for men

while the lower rate in FFSPPO offices was significant only for female patients.

Among patients with no chronic medical conditions, office visits to all four insurance

categories of primary care providers had statistically similar probabilities of including

a therapeutic or preventive service.

With respect to diagnostic and screening services, office visits to providers who

accepted both HMOPPO or pure HMO plans had a greater likelihood of including

these services than did visits to physicians who accepted only FFS insurance plans,

even when controlling for the patient's health status and demographic characteristics.

In the case of therapeutic and preventive services, however, only visits to pure HMO

offices had a significantly higher likelihood of receiving such a service than visits to

pure FFS offices. For both diagnosticiscreening services and therapeuticipreventive

services, the presence of a chronic medical condition exerted a much stronger

influence on the likelihood that a patient would receive either kind of service than did

the types of insurance coverage accepted by the provider

The finding that female patients were more likely than male patients to have

received diagnostic or screening services regardless of the health care provider's

insurance aftiliation most likely reflects female patients' greater frequency of visits for

exams and check-ups. More than 63% of all office visits in 1995 were made by

female patients, and a subset of these visits - such as those for mammograms or pap

smears -a h y s involve a test or exam. The greater probability of receiving either

diagnosticiscreening services or therapeuticlpreventive services in an HMO-only

office compared to a physician office that accepts only FFS insurance plans could be

the result of several characteristics of HMOs. Health maintenance organizations have

historically emphasized preventive care as a way of promoting good health and

potentially reducing health care expenses over the long term. HMOs also frequently

screen patients with phone consultations before making an appointment for an office

visit. If effective, this would result in a greater proportion of patients who ultimately

see a physician in a pure HMO being in need of medical services compared with

patients in purely FFS physician offices. Finally, an emphasis on primary care is

among the hndamental principles of HMOs, and HMOs are staffed mainly by

primary-care physicians. Consequently, it may be that some services performed in the

offices of primary care physicians in HMOs are referred to specialists in FFS office

settings.

Differences in the frequency with which diagnostic or therapeutic services are

performed by any given group of physicians on any given group of patients do not

necessarily mean that the rates at which these services are performed in one case or

the other is inappropriate. However, while this analysis showed that a patient's health

status has a much stronger relationship to the likelihood of receiving services during

an office visit than does the provider's insurance affiliations, these too appear to

influence the likelihood of some kinds of services be administered. Differences in the

rates at which diagnostic and therapeutic services are performed among male and

female patients and among office visits to physicians who participate in HMOs or

PPOs or do not participate in any kind of managed care naturally raise the question

of why these differences occur.

Previous research has shown that patients enrolled in HMOs typically have a

higher number of physician office visits and fewer inpatient admissions in a given

period of time than non-HMO patients, but that they also have less access to

specialists and use fewer complex tests and procedures. This analysis presents

evidence that, even when controlling for the patient's health status, diagnostic and

screening services are most likely to be provided during office visits to primary care

physicians in HMO or HMOPPO practices, and that therapeutic and preventive

services are most likely to be performed during ofice visits to HMO-only physician

practices.

Appendix I: Standard Errors of Estimates

Table A l . Standard Errors for Estimates Displayed in Table 1

Patient age, sex, race;

physician location

Under age 18

Ages 18 to 34

Ages 35 to 54

Ages 55 to 64

Female

Male

White

Black or other

Has chronic condition(s)

No chronic conditions

Northeast

Midwest

South

West

Urban

Rural

Physician:

Has seen patient before

Has not seen patient before

Total

Visits

(000s)

FFS only

Standard error (+/-)

FFS or

HMO or

PPO

PPO

HMO only

Table A2. Standard Errors for Estimates

Displayed in Table 2

Were dzagnostzc servzcer performed7

Type of insurance

Total visits

YesINo

accepted by physician

(000s)

Standard error (+I-)

FFS only

HMO only

FFS or PPO

HMO or PPO

Total

16,848

9,662

27,075

114,703

168,288

2.0%

2.3%

1.7%

0.8%

0.7%

People Who Had No Chronic Medical Conditions

Were diagnostzc services performed?

Type of insurance

Total visits

YeslNo

accepted by physician

(000s)

Standard error (+I-)

FFS only

HMO only

FFS or PPO

HMO or PPO

Total

11,459

7,115

21,915

91,365

131,854

2.6%

2.9%

2.0%

1.0%

0.8%

Table A3. Standard Errors for Estimates Displayed

in Table 3

Female Patients

Were diagnostic or screening services performed?

Type of insurance

Total

Yes /No

accepted by physician

visits (000s) Standard error (+I-)

2. 7%

9,408

FFS only

2.8%

6,194

HMO only

2.1%

16,738

FFS or PPO

1.0%

74,602

HMO or PPO

106,942

0.8%

Total

Male Patients

Were diagnostic or screening services performed?

Type of insurance

Total

Yes /No

accepted by physician

visits (000s) Standard error (+I-)

7,440

3.0%

FFS only

3,469

4.0%

HMO only

10,337

3.0%

FFS or PPO

1.5%

40,101

HMO or PPO

1 2%

61,347

Total

Table A4. Standard Errors for Estimates Displayed

in Table 4

All Privately Insured People Under Age 65

Were therapeutic or preventive seivjces performed?

Type of insurance

Total

YeslNo

accepted by physician

visits (000s) Standard error (+/-)

FFS only

16,848

2.3%

HMO only

9,662

3.2%

FFS or PPO

27,075

1.8%

HMO or PPO

114,703

0.9%

Total

168,288

0. 7%

People with No Chronic Medical Conditions

Were therapeutic or preventive sewices performed?

Type of insurance

Total

YeslNo

accepted by physician

visits (000s) Standard error (+/-)

FFS only

11,459

2.7%

HMO only

7,115

3.6%

FFS or PPO

1.9%

21,915

HMO or PPO

91,365

1.0%

Total

131,854

0.8%

Table A5. Standard Errors for Estimates

Displayed in Table 5

Female Patients

Were therapeutic or preventive sewzces per$ormed?

Type of insurance

Total visits

YesINo

accepted by physician

(000s)

Standard error (+/-)

FFS only

9,408

3.2%

HMO only

3.9%

6,194

FFS or PPO

16,738

2.2%

HMO or PPO

74,602

1.1%

Total

106,942

0.9%

Mate Patients

Were therapeutic or preventive services performed?

Type of insurance

Total visits

YeslNo

accepted by physician

(000s)

Standard error (+/-)

FFS only

7,440

3.5%

HMO only

3,469

5.3%

FFS or PPO

2.9%

10,337

HMO or PPO

40,101

1.5%

Total

61,347

1.2%

Appendix 11: Medical Services and Medical Conditions

The National Ambulatoly Medical Care Survey (NAMCS) defines three

categories of diagnostic and screening services: exams, tests, and imaging.

Exams, Tests, and Imaging

Breast exam

Pelvic exam

Rectal exam

Visual acuity exam

Mental status exam

Other exam

Blood pressure test

Urinalysis

TB skin test

Blood lead level test

Cholesterol measure

Prostate screening antigen (PSA) test

HIV serology

Other blood test

Other test

X-ray

Computerized axial tomography (CAT scan)

Magnetic resonance imaging (MRI)

Ultrasound

Other imaging

Other diagnostic or screening service

The NAMCS defines two categories of therapeutic and preventive services:

counselingleducation and other therapy.

CounselinglEdueation

Diet

Exercise

Weight reduction

Cholesterol reduction

HIV transmission

Tobacco uselexposure

Growthidevelopment

Mental health

Other counseling

Other Therapy

Psychotherapy

Corrective lenses

Physiotherapy

Other counseling

References

Clement, D.G., S.M.Retchin, R.S. Brown, and M.H. Stegall. Access and Outcomes

of Elderly Patients Enrolled in Manased Care. Journal of the American Medical

Association, v. 271, no. 19, May 18, 1994.

CRS Report 97-482, Managed Health Care: The Use of Firzancial Incentiiles, by

Beth C. Fuchs and Jason S. Lee.

CRS Report 97-913, ManagedHealth care: A Primer, by Jason S. Lee.

Eddy, D.M. Balancing Cost and Quality in Fee-for-Service Versus Managed Care.

Health Affnic~,v. 16, no. 3, MayIJune 1997.

Fama, T., P.D. Fox, and L.A. White. Can HMOs Care for the Chronically Ill?

Health Affairs, v. 14, spring, 1995.

Gabel, J. Ten Ways HMOs Have Changed During the 1990s. Health Affnirs, v. 16,

no. 3, MayIJune 1997.

Gazmararian, J.A. and J.P. Koplan. Length of Stay After Delivery: Managed Care

Versus Fee-for-Service. Health ASfairs, v. 15, no. 4, winter, 1996.

Hellinger, F.J Selection Bias in HMOs and PPOs: A Review of the Evidence.

Inquiry, v. 32, summer, 1995.

Hoechst-Marion-Roussel. Managed Care Dlgest Series: HMO-PPO Digest 1996.

Kansas City, MO, 1996.

KPMG Peat Marwick. The Impact ofManaged Care on US. Markets. Washington,

DC, 1996.

Mark, T. and C. Mueller. Access to Care in HMOs and Traditional insurance Plans.

Health Affairs, v. 15, no. 4, winter, 1996.

Miller, R.H. and H.S. LuR. Does Managed Care Lead to Better or Worse Quality of

Care? Health Affairs, v. 16, no. 5, SeptemberIOctober 1997.

Miller, R.H. and H.S. Luft. Managed Care Plan Performance Since 1980. Journal

of the American Medical Association, v. 271, no. 19, May 18, 1994.

Robinson, J.C. Decline in Hospital Utilization and Cost Inflation Under Managed

Care in California. Journal of the American Medical Association, v. 276, no.

13, October 2, 1996.

Safran, D.G., A.R. Tarlov, and W.H. Rogers. Primary Care Performance in Fee-forService and Prepaid Health Care Systems. Journal ofthe Anzerican Medical

Association, v. 271, no. 20, May 25, 1994.

Taylor, A.K., K.M. Beauregard, and J.P. Vistnes. Who Belongs to HMOs: A

Comparison of Fee-for-Service Versus HMO Enrollees. Medical Cure Research

andReiliew, v 52, no. 3, September 1995.

Ware, J.E., M.S. Bayliss, W.H. Rogers, M. Kosinski, and A.R. Tarlov. Differences

in 4-Year Health Outcomes for Elderly and Poor, Chronically I11 Patients Treated

in HMO and Fee-for-Service Systems. Journal of the American Medical

Association, v. 276, no. 13, October 2, 1996.

Weiner, J.P., A. Lyles, D.M. Steinwachs, and K.C. Hall. Impact of Managed Care

on Prescription Drug Use. Health Affairs, v. 10, spring, 1991.

Wholey, D.R., J.B. Christianson, J. Enberg, and C. Bryce. HMO Market Structure

and Performance. Health Affairs, v. 16, no. 6, NovemberiDecember 1997.

Yelin, E.H., L.A. Criswell, and P.G. Feigenbaum. Health Care Utilization and

Outcomes Among Persons with Rheumatoid Arthritis in Fee-for-Service and

prepaid Group Practice Settings. Journal of the American Medical Association,

v. 276, no. 13, October 2, 1996.