The effect of coordinated, multidisciplinary ambulatory care on service use, charges, quality of care and patient satisfaction in the elderly

.Journal of Community Health Vol. 18, No. 2, April 1993

THE EFFECT OF COORDINATED, MULTIDISCIPLINARY AMBULATORY

CARE ON SERVICE USE, CHARGES, QUALITY OF CARE AND PATIENT SATISFACTION IN THE ELDERLY

Laura-Mae Baldwin, MD, MPH; Thomas S. Inui, ScM, MD; and Sally Stenkamp, MD

ABSTRACT: This study evaluated a muhidisciplinary care center, the Pike Market Clinic (PMC), whose physicians provide and coordinate inpatient and outpatient care fbr downtown low-income elderly in Seattle. We interviewed users of PMC and their near-neighbors with a 206 item questionnaire to compare their medical and social service use, quality of care, and satisfaction. We then estimated mean annual charges/person for inpatient, outpatient and emergency room services in the two groups. Demographic and health status characteristics were similar in the two groups. PMC patients made significantly more annual visits than neighbors to their primary physicians. Visits to non-primary physicians oc- curred at the same rate in both groups, but PMC patients were referred more often by their primary physicians. Both emergency room and inpatient use were higher in the neighbor group. Social services were used at the same rate by PMC patients and neighbors. Various indices suggested that quality of care and satisfaction were comparable or superior among PMC patients. Using utilization data, we estimated that neighbors generated charges over $1000/person/year greater than PMC patients. Coor- dination by PMC providers rather than the availability of muhidisciplinary services may be largely responsible for utilization differences between PMC patients and their neighbors.

INTRODUCTION

The growing number of aged in our country is no longer a pro- jection--i t has become a reality. The number of persons aged 65 and

Laura-Mae Baldwin is Assistant Professor of Family Medicine at the University of Wash- ington, Seattle, Washington. Thomas S. Inui is Professor and (;hair of Ambulatory (;are and Pre- vention at Harvard Medical School, Boston, Massachusetts. Sally Stenkamp is in practice at Kaiser Permanente in Riverside, California.

This research was conducted while the first author was a Robert Wood Johnson Clinical Scholar at the University of Washington. The opinions and conclusions in the text are those of the author and do not necessarily represent the views of The Robert Wood Johnson Foundation. The authors would like to thank Marilyn Bergner, PhD for her review of" this work, Betsy Lieberman tot her administrative support at the Pike Market Clinic, Beth Lindeman, MD for her assistance in interviewing and the many elderly who opened their homes to the interviewers.

Requests 1or reprints should be addressed to: Dr. Laura-Mae Baldwin, Department of Fanfily Medicine, RF-30, University of Washington School of Medicine, Seattle, Washington 98195.

© 1993 Human Sciences Press, Inc. 95

96 JOURNAL OF COMMUNITY HEALTH

older increased 86% between 1960 and 1989, from 16.7 million to 31 million, and is slated to increase an additional 3.5 million by the year 2000. ~ Because of the greater morbidity and disability of the elderly, the increas- ing size of this group has had profound effects on economic, social, and health service institutions. For example, although 12% of the U.S. population was considered elderly in 1986, 33% of the health care dollar was spent on this group, primarily in the hospital and nursing home settings5

Health planners and providers have sought alternatives to the costly, traditional manner in which health care services are provided to the elderly? A multidisciplinary approach to providing care to this group has been developed and promoted over the past decade. A plethora of programs employing physicians, nurses, social workers, and other support personnel has attempted to target the elderly either eligible or at risk for institutionalization, and to provide them with a com- prehensive set of outpatient and in-home services? -6 These programs hope to maintain the elderly in their homes, thereby avoiding expensive inpatient care and long-term institutionalization while improving outcomes such as functional status and mortality.

Indeed, several studies have demonstrated such outcomes. Weiss and Sklar conducted a randomized clinical trial with 338 elderly persons with health and social service needs that diminished their chances of living independently. 7 The experimental group, which received a functional needs assessment followed by a coordinated set of social and health services either in adult day centers or the home, showed less de- terioration in functional status, 17% lower overall costs, and a 12% re- duction in acute inpatient admissions compared to the control group. Two quasi-experimental studies used multidisciplinary teams to provide a full range of community-based health and social services to a total of 953 subjects? ''~' These studies found similar decreases in inpatient use, with the experimental groups spending 29% and 33% fewer days in the inpatient setting. One of the two quasi-experimental studies examined costs, and found them to be 12% lower in the intervention group.

Other studies have reported less optimistic results. Hedrick and Inui have reviewed 12 experimental and quasi-experimental studies of home care targeting chronically ill groups, and found that in general, experimental groups demonstrated the same or increased inpatient and outpatient service use and costs of care.L° A second review by Weissert, Cready and Pawelak of 27 home and community-based long-term care studies found that the majority of these programs raised health care service use and costs. L~

Certain patient subgroups did experience benefits from these long-term care programs, however. The elderly who were not severely

Laura-Mae Baldwin et al. 97

disabled and who had only moderate unmet needs had lower hospital use, while the younger elderly with less disability and better social support were more likely to experience improved survival, mental functioning, and mobility. This less disabled subgroup represents a greater proport ion of the elderly in our country than the group at risk fl)r institutionalization. While this less disabled group may not require the intensive care that these home and community-based long term care programs offer, they may receive similar benefit f rom access social services on an intermit tent or less

One vulnerable segment of this income, urban elderly, who often live

to a broad range of health and intense basis. less disabled group is the low- alone and may depend more

heavily on social services. This study evaluates an ambulatory care source, the Pike Market Clinic, which provides a well-coordinated, multidisciplinary set o1 health and social services to a low income, urban, elderly population in Seattle, Washington. The care model used by the Pike Market Clinic expands beyond the traditional medical and social services offered at most comlnunity health centers. If successful, this model could serve as a prototype for other clinical centers seeking to offer the highest quality care to the elderly.

We have evaluated this ambulatory (:are model by contrasting the medical care utilization patterns of two g roups - - r egu la r patients of the Pike Market Clinic (PMC) and their near-neighbors who chose other medical care sources. We assessed whether the availability of coordinated or multidisciplinary services at the Pike Market Clinic affected the inpatient and outpatient service use, social service use, quality of care, patient satisfaction, or estimated charges for care among PMC patient compared to their neighbors.

Study Sett ing

The Pike Market Clinic is a community-based ambulatory health care center that was established in 1978 in response to the results of a communi ty needs assessment (Hurley C: Communi ty Needs Assessment Questionnaire, October, 1977). This survey of 185 low-income elderly in downtown Seattle found that 37% of this group reported external bar- riers to care, such as lack of transportation and inability to pay, that had prohibited them from obtaining desired medical attention dur ing the previous 12 months. Since that time, the clinic has worked to provide a coordinated, multidisciplinary set of services tailored to the needs of an elderly community.

The services offered by the clinic at the time of the study included:


1. primary medical care provided by two full-time internists and a part-time physician's assistant who were on-call 24 hours a day;

2. home health nursing care coordinated with the patient's medical care through ongoing discussion and conferences among the providers;

3. mental health and alcoholism counseling in an individual or group setting with either a part-time psychiatrist or one of two mental health counselors;

4. full-time social service referral in areas such as transportation, housing, home helper services and meals;

5. health education classes at the clinic or a nearby senior center on common geriatric problems such as hypertension, depression and diabetes;

6. a pharmacy with at-cost medications; 7. laboratory services; 8. physical and massage therapy; and 9. podiatry (Annual Report, Pike Market Community Clinic,

1985).

The clinic is located in downtown Seattle, central to a number of subsidized housing units, and close to bus service. There is a sliding fee scale, and Medicare assignment is accepted. Forty-six percent of the clinic's funding came from local, state and federal governmental sources, 31% from individual donations and grants, and 23% from patient revenue.

The clinic strongly requests that its patients use the clinic as their only primary care source by recommending that patients with another primary physician choose between the two for their regular care. Inpa- tient service, specialty consultations, and radiology examinations are provided almost exclusively at one inpatient facility, Pacific Medical Center, which also offers a sliding fee scale. During hospitalization, a Pike Market Clinic physician serves as the attending physician. The Pike Market Senior Center is located in the same building as the clinic. While administered separately, it is closely tied to the clinic, and offers an additional system of social support to this population.

M E T H O D S

We identified two study populations--users of the Pike Market Clinic and their near-neighbors who used non-PMC health care providers. To be eligible for the study, members of both study groups needed to 1) be 60 or older,


2) live within the catchment area of the Pike Market Clinic, and 3) identify a regular medical provider whom they had seen within the six months prior to their participation in the study. We excluded elderly who did not speak English or were so demented that they could not answer questions reliably.

We identified Pike Market Clinic patients who were potentially eligible for the study from the clinic's computer system. In accordance with the Univer- sity of Washington Human Subjects Committee requirements, a clinic staff member approached these patients to ask permission for an interviewer to contact them. If permission was granted, the interviewer explained the research effort and obtained consent for an interview. The majority of Pike Market Clinic patients participating in the study lived in one of eleven hotels or apart- ment buildings in downtown Seattle. The remainder lived in smaller apart- ments or hotels, and single dwellings in the downtown area. We sought the neighbor population from the eleven buildings with the greatest number of Pike Market Clinic patients. Managers identified the room numbers of all residents 60 and older in these buildings, and we sent letters describing the study to the residents of these rooms. Following the letter, three trained interviewers attempted to locate the residents in their rooms to identify neighbors who were potentially eligible for the study.

We developed a 206 item survey which included questions about use of medical and social services, satisfaction with care, and the care received for certain medical conditions. Sections of the survey were adapted from several sources: the utilization portion of the National Opinion Research Center Sur- vey of Access to Medical Care, '~ L. Branch's social service needs questions from the Massachusetts Longitudinal Survey of the Elderly, ':~ and J. Ware's Short Form Patient Satisfaction Survey. ~4 One of the three interviewers administered this one-hour questionnaire to members of the study groups, usually in the subjects' homes. We conducted interviews from .July 1 through September 30, 1985.

We calculated estimated mean annual charges per person for inpatient, outpatient and emergency room services in the two groups using utilization data from the survey and charge data from several sources. The Health Care Financing Administration provided the mean inpatient hospitalization charge for a Washington State Medicare recipient in 1985. Schneeweiss and his col- leagues provided unpublished outpatient and emergency room charge data for patients 65 and older who received care in 1986 at the Family Medical Center, University Hospital in Seattle, Washington. These charges cover the provider's component of care only, excluding laboratory or radiology fees. These data were originally collected as part of another study. '~

We used chi-square analysis to compare categorical variables and t-tests to compare continuous variables between the PMC and neighbor groups. Mul- tivariate regression techniques controlling tor gender, disability days and self- reported health status were performed when examining the effect of source of care on utilization. Since no differences were found between the bivariate and multivariate analyses, the bivariate findings are reported here.


R E S U L T S

One hundred and fifty Pike Market Clinic patients and 263 of their neighbors were identified as potentially eligible for the study. One hundred and nine (73%) of the Pike Market Clinic patients were interviewed, 20 (13%) refused, and 21 (14%) were never found. Of the neighbors, 164 (62%) were interviewed, 77 (29%) refused, and 22 (9%) were never found.

The demographic characteristics of the Pike Market Clinic and neighbor groups are strikingly similar. The majority of both groups were white (87% and 86% of the Pike Market Clinic patients and neighbors, respectively), lived alone (92% and 93%), and had annual incomes below 150% of the poverty level (82% and 88%). About one fourth of the subjects had some college or vocational training after completing high school. The mean age of both PMC patients and their neighbors was about 71 years. Gender and insurance status differed between the two groups. A significantly greater proportion of Pike Market Clinic patients were male (67%), compared to the neighbors (51%, p=.01). Pike Market Clinic also sees a higher percentage of patients with no insurance (10%), compared to the neighbor group (4%, p= .04).

We used three measures of health status to examine the compa- rability of the two groups. Self-perceived health status was measured with the question, "How would you say your health is now--excellent, good, fair, or poor?" A second measure examined the number of days in the past two months that a participant had been in bed all or most of the day because of illness or a health condition. This did not include days in the hospital or a nursing home. Third, we looked at the average number of prescription medications patients in each group were taking. Although there were no statistically significant differences between the two groups, a slightly greater percent of Pike Market Clinic patients (57%) reported fair or poor health status, compared to their neighbors (54%, p=0.47). Alternately, a greater percent of the neighbors had more than a week of disability days in the past two months (9%), compared with PMC patients (3%, p = .06). The mean number of prescrip- tions taken by patients in each group was equivalent, 2.4 (p = .95).

The neighbors sought their health care from a wide variety of providers. The largest group, about 37%, received care in a private physician's office. Hospital outpatient departments accounted for an additional 20% of the care, with the Veterans Administration facilities and Health Mainte- nance Organizations third (16%) and fourth (12%) in frequency.

Neighbors used sources of care differently than patients who

Laura-Mae B a l d w i n et al. i 01

used the Pike Market Clinic. In the two months prior to the interview, Pike Market Clinic patients made an average of 3.0 visits, whereas neighbors made only 1.7 visits (p = .03). The average number of visits to physicians other than the regular, primary care provider in the six months prior to the interview was essentially the same in both groups, 1.5 for Pike Market Clinic patients and 1.6 for neighbors, ahhough the types of providers and the way in which they were used was quite different. Neighbors (15%) used other primary care providers in addition to their "regular" providers more often than Pike Market Clinic patients (3%). Medical subspecialists were used at the same rate in the two groups (20%), whereas referral to surgeons was more frequent in the PMC group (34% versus 22%). The way in which patients were referred to the "non-regular" physicians in the two groups is also of interest. Pike Market Clinic patients were referred by their primary provider 49% of the time, neighbors only 36%.

Concomitant with these differences in outpatient use were marked differences in the use of institutional services. Neighbors used an average of 37 emergency room visits per 100 persons in six months, compared to 11 visits made dur ing the same time by PMC patients (p=.002). Neighbors, with an average of 32 hospitalizations per 100 persons in six months, also used inpatient hospital services more fre- quently than PMC patients, who had 16 hospitalizations per 100 persons dur ing the same six month period (p = .07).

As expected, both PMC patients and their neighbors who reported fair or poor health used more health services than those who reported excellent or good health (Table 1). The differences noted pre- viously between the PMC and neighbor groups in utilization of outpatient and inpatient services were most striking for patients report ing fair or poor health status. PMC patients and their neighbors in excellent or good health made similar use of their primary care providers, other providers and hospital services. PMC patients in excellent or good health had significantly fewer emergency room visits, however. PMC patients who reported lair or poor heahh visited their primary care provider twice as often as neighbors in fair or poor health. They were hos- pitalized and seen in emergency rooms much less than neighbors with similar health status.

Although the use of physician services was quite different in the two groups, we found no significant differences between the groups in the use of mental health counseling, home health nursing care, meals in the home, and home helper services in the six months prior to the interview.

102 J O U R N A L O F C O M M U N I T Y H E A L T H

TABLE 1

Medical Service Utilization Stratified by Health Status*

Excellent~Good Health Fair~Poor Health PMC Neighbors PMC Neighbors

N=31 N=61 N=48 N=68

Primary Care MD 1.70 1.50 3.80** 1.80 Visits/2 months Other MD Visits/6 1.50 1.60 1.50 1.50 months Hospitalizations/6 0.13 0.10 0.19** 0.51 months Emergency Room 0.03** 0.18 0.17** 0.54 Visits/6 months

*Ns are smaller because samples include only those elderly using current health care source for greater than 1 year.

**p<.05 by chi-square for all paired comparisons (underlined)

We examined both quality of care and patient satisfaction in our two study groups. Using strategies employed by Heller et al. ~6 to examine quality of care in the elderly, we compared t reatment of common geriatric problems such as hypertension, depression, angina, and hear- ing loss and use of preventive measures, including influenza vaccine, guaiac testing, and breast cancer screening. The prevalence of disease was similarly high in both groups. In the treatment of common geriatric problems and in the use of preventive measures, the Pike Market Clinic provided the same or better care in all areas except one - -p rov ide r performed breast examination within the last year.

Patient satisfaction scores, while high for both groups, were con- sistently higher for the Pike Market Clinic patients than their neighbors (Table 2). The Pike Market Clinic patients were significantly more satis- fied than their neighbors in the area of convenience of care.

As expected from the utilization differences between the two groups, estimated total annual charges per person for care were higher in the neighbor group than in the Pike Market Clinic group (Table 3). On average, we estimate that neighbors generated charges over $1000 per person per year greater than Pike Market Clinic patients. Although Pike Market Clinic patients spent 1.4 times more on outpatient clinic services, their dramatically decreased used of emergency room and in-


TABLE 2

Patient Satisfaction

No. Items PMC Neighbors (Mean Scores)

N = IOI* N = 1 3 8 "

P

Accessibility 5 17.74 17.14 .17 (Maximum = 25) Convenience 2 7.95 7.35 .004 (Maximum = 10) Humaneness 8 29.28 28.96 .64 (Maximum = 40) MD Quality 7 24.84 23.79 .06 (Maximum = 35) General Satis- 4 14.41 14.31 .80 faction (Maxi- mum = 20)

*Number varies slightly fbr each item according to missing data.

TABLE 3

Estimated Charges Per Person Per Year (Dollars)

PMC Neighbors

Inpatient 1323.00 2645.00 Outpatient 1092.00 778.00 Emergency Room 43.00 144.00 Total 2458.00 3567.00

patient services accounted for the lower overall charges compared to the neighbors.

DISCUSSION

The two groups whose medical care experience was examined in this study, elderly Pike Market Clinic patients and their near-neighbors, were remarkably similar in their health status and demographic charac-


teristics. Only two significant differences were f i )und- - the Pike Market Clinic serves a larger proport ion of male patients and patients without insurance. This may be explained by two features: 1) the close referral relationship between the clinic and the senior center, which serves a predominant ly male population, and 2) Pike Market Clinic's policy to accept patients regardless of their ability to pay. In general, however, the Pike Market Clinic attracted patients much like those seeking care from other sources.

Despite the similarities in health status and demographics of the two groups, their service use pattern differed strikingly. First, Pike Mar- ket Clinic patients appear to have substituted outpatient for emergency room and hospital services. This result is consistent with the findings of other studies examining the effects of community health centers on hospital-based service use. ~v-'-'' Second, near-neighbors used other primary care physicians in addition to their "regular" providers more often than Pike Market Clinic patients. In addition, they were refer red to other consuhant and specialty physicians by their "regular" provider less fre- quently than Pike Market Clinic patients. These utilization differences resulted in estimated charges that were 31% lower for the PMC group than for their neighbors. These data suggest that the primary care providers at the Pike Market Clinic coordinated a more parsimonious, cost- effective set of medical services for their patients. •

Coordinat ion of care aside, there is no evidence that the multidisciplinary nature of the Pike Market Clinic services can explain these differences. Despite offering a wide range of social and health services at the Pike Market Clinic, there were no significant differences in the use of these services between the two study groups. Compared to the elderly surveyed by Branch, ~ the percent of patients in both the Pike Market Clinic and neighbor groups who used these multidisciplinary services was quite high. The social service needs section of the study questionnaire, not repor ted in detail here, shows similar percents of both groups with unmet social service needs. Perhaps the maximum number of patients able to afford the services offered was reached in both groups, making the limiting factor the availability of funds for the services, ra ther than the availability of the services themselves.

The differences in service use appear to have been a function of the coordination of care, ra ther than the availability of a multidisciplinary set of services. Pike Market Clinic providers acted as "case managers" for their patients, and influenced both the volume and types of medical services that their patients used without compromising the quality of care they received or their satisfaction with care. Evaluations of


several managed care systems have arrived at similar results. ~:~-''~' Weil's determination of costs to Medicare of seven prepaid group practices found that the prepaid plans had lower hospital and emergency room costs when compared to the open market. ='~ Although home health and outpatient costs were higher, the net effect was lower Medicare expenditures in the prepaid group practices. While Well was able to control for age and sex differences between the prepaid and control groups, he did not know whether there were health status differences between the two groups that might account for the cost differences. Manning et al. '-'~' overcame this potential source of bias with their randomized controlled trial, which showed comparable decreases in hospitalization and expen- diture rates in a group using a prepaid group practice rather than fee- for-service practitioners. Manning's work excluded elderly subjects from their study population, however, making generalization from this group problematic.

Two more recent studies examining the effect of managed care systems on use of health services by the elderly have had less consistent findings. Wan's evaluation of a managed care demonstrat ion program sponsored by the Health Care Financing Administration found few effects on the use of outpatient and inpatient services by elderly eligible fi)r both Medicaid and Medicare. '-'~ Thomas and Kelman compared health care utilization among urban elderly in a hospital-based health maintenance organization (HMO), a preferred provider organization (PPO) and the fee-for-service system. '-'7 After adjusting for health and socioeconomic characteristics of the plan's enrollees, the authors did not demonstrate the expected lower rates of hospitalization and shorter lengths of stay in all three managed care systems. Only the PPO plan members had lower use of hospital services. There was no concomitant higher use of outpatient services in this group. Unlike the Pike Market Clinic patients, the elderly patients in these other managed care systems did not increase their outpatient service use substantially. Frequent outpatient contact and support may have been the missing element in these managed care programs.

Evaluations of Medicare's prospective hospital payment system (DRGs) provide additional support for the finding that outpatient care can be substituted for inpatient care for the elderly, potentially resulting in cost savings. '-'~~' While the increased outpatient service use associated with DRGs has primarily been a reaction to the shortened inpatient lengths of stay, hospital admission rates have also fallen, suggesting that outpatient services can be used in lieu of some inpatient care.

There are marked differences between the two groups examined

106 J O U R N A L O F C O M M U N I T Y H E A L T H

in this study and the general elderly population in the U.S. Branch's 1976 probability survey of elderly in Massachusetts shows that only 39% of elderly living alone reported their health as fair or poor, in contrast to the approximately 55% of elderly in our sample. '~ In 1985, people 65 and older made an average of 4.9 office visits per person per year, making the neighbor and Pike Market Clinic groups between 2.1 and 3.7 times more likely to visit a physician? ° The hospitalization rate for the neighbors follows a similar t r e n d - - t h e six-month rate, 0.36 hospitalizations/person/year, nears the 1985 West Coast annual rate of 0.40. 3' The low socioeconomic status of our study groups, as well as the eligi- bility requi rement that all subjects identify a regular provider they had seen in the six months prior to the interview, may explain the poorer health and greater service use of our two groups.

The Health Care Financing Agency reports that the average 1985 Medicare payment per enrollee in Washington State totaled $2059. As we would expect of identified users of care who report rela- tively low self-perceived health status, the neighbor group's estimated charges were about $1500 higher than those reported by HCFA. Pike Market Clinic patients' estimated charges, on the other hand, were only $400 more than Medicare's yearly average, a potential savings of $1100 per person when compared to their neighbors.

The limitations of this study prohibit broad generalizations. First, the patients were able to select their own source of medical care, which introduces self-selection bias. There may be unmeasured variables that vary between the two groups and account for the differences in use of and charges for services. Our findings were verified, however, by multivariate analyses controlling for gender, self-reported health status and disability days. Another limitation of the study was the lack of blinding of the interviewers to the care source of the subjects. Lastly, only three practitioners provide medical care at Pike Market Clinic, making it difficult to separate the effects of practitioner style from the effects of the system itself. Despite these limitations, this study does suggest that a coordinated system of primary care can substantially decrease the use of institutional services and the charges for medical services in an elderly population while maintaining quality of care and patient satisfaction. Further work examining other practitioners and settings will be needed to confirm the findings of this study and to support the generalizability of its results.

REFERENCES

1. U.S. Bureau of the Census: Statistical Abstracts of the United States: 1991 (111th edition.) Wash- ington, DC, 1991. Pp. 13, 15.


2. U.S. Senate Special Committee on Aging: Aging America: TrencL~ and Projectiort~ (I987-88 edition). U.S. Dept. of Health and Human Services. P. 125.

3. Health and Public Policy Committee, American College of Physicians, Long term care of the elderly. Ann Intern Med 100:760-763, t984.

4. Zawadski, RT, The long term care demonstration projects: What are they and why they came into being. Home Health Care Services Quarterly 4: 5-19, 1985.

5. Kodner, DL, Mossey, W, Dapello, RDL, New York's "Nursing Home Without Walls": A provider-based community care program fi)r the elderly. Home Health Care Services Quarterly 4:107-126, 1985.

6. Eggert, GM, Bowlyon, JE, Nichols, CW, Gaining control of the hmg term care system: First returns from the ACCESS experiment. Gerontologist 20:356-363, 1980.

7. Weiss, LJ, Sklar, BW, Project OPEN: A hospital-based long-term care demonstration program for the chronically ill elderly. Home Health Care Services Quarterly 4:127-145, 1985.

8. Ansak, M-L, Zawadski, RT, On Lok CCODA: A consnlidaled model, ttome Health Care Services Quarterkv 4:147-170, 1985.

9. Brill, RS, Horowitz, A, The New York City Home (_;are Proiect: A demonstration in coordination of health and social services. Home Health Care Se~v~ices Quarterly 4: 91-106, 1985.

10. Hedrick, SC, Inui, TS, The effectiveness and cost of home care: an information svnthesis. Health Serv Res 20: 851-880, 1986.

11. Weissert, WG, Cready, CM, Pawelak, JE, The past and fiature of home- and community-based long-term care. Milbank Q 66:3(19-388, 1988.

12. Aday, LA, Andersen, R, Fleming, GV: A National Survey, o/Access to Medical Care. Center for Health Administration Studies, Univ. of Chicago, 1977.

13. Branch, LG, Fowler, FJ: The health care needs ~the elderly and chronically disabled in Massachu.~ett.~. Survey Research Program Monograph, 1975.

14. Ware, JE, Snyder, MK, Wright, WR: Scoring instructions.[or short forms o[ the patient satisfaction questionnaire. Health Insurance Study. U.S. Dept. of tleahh, Education, and Weffare, Wash- ington, D.C.

15. Schneeweiss, R, Ellsbury, K, Hart, LG, Geyman, JP, The economic impact and multi- plier effect of a family practice clinic on an academic medical center..lAMA 262:370-375, 1989.

16. Heller, TA, Larson, EB and Lo(;erti), JP, Quality of ambulatory care of the elderly: An analysis of five conditions. J Am Geriatr Soc 32:782-788, 1984.

17. Bellin, SS, Geiger, H J, Gibson, CD, Impact of ambulatory-heahh-care services on the demand for hospital beds. N EnglJ Med 280:808-812, 1969.

18. Hochheiser, LI, Woodward, K, Charney, E, Efti~ct of the neighborhood health center on the use of pediatric emergency departments in Rochester, New York. N Engl.] Med 285:148-152, 1971.

19. Master, RJ, Feltin, M, Jainchill, J et al., A continunm of care fi)r the inner city: Assessment of its benefits for Boston's elderly and high risk populations. N EnglJ Med 302: 1434-1440, 1980.

20. Okada, LM, Wan, TTH, hnpact of community heahh centers and Medicaid on the use of health services. Public Health Rep 95:520-534, 1980.

21. Deprez, RD, Pennell, B-E, Libby, MA, The substitutability of outpatient primary care in rural community health centers for inpatient hospital care. Health Serv Res 22:207-233, 1987.

22. Branch, LG: Understanding the health and social service needs of people over age 65. (;enter l~)r Survey Research, Univ. of Mass., and The Joint Center for Urban Studies, M.I.F'. and Har- vard Univ., 1977.

23. Luft, HS, How do heahh-maintenance organizations achieve their "savings"? N EngIJ Med 298:1336-1343, 1978.

24. Well, PA, Comparative costs to the Medicare program of seven prepaid group practices and controls. Milbank Q Summer:339-365, 1976.

25. Manning, WG, Leibowitz, A, Goldberg, GA, Rogers, WH, Newhouse, JP, A controlled trial ot" the effect of a prepaid group practice on the use of services. N Engl] Med 310:1505-1510, 1984.

26. Wan, TTH, The effect of managed care on health services use by dually eligible elders. Med Care 27:983-1001, 1989.

27. Thomas, C, Kehnan, HR, Health services use among elderly under alternative heahh service delivery systems. J Community Health 15:77-92, 1990.

28. Menke, T, Impacts of PPS on Medicare Part B expenditures and utilization for hospital epi- sodes of care. Inquiry 27:114-126, 1991t.


29. lglehart, JK, Early experience with prospective payment of hospitals. N Engl.] Med 314: 1460- 1464, 1986.

30. National Center for Health Statistics: Health, United States, 1986. DHHS Pub. No. (PHS) 87- 1232. Public Health Service.

31. National Center |br Health Statistics, E.J. (;raves: Utilization of short-stay hospitals, United States, 1985. Vital and Health Statistics. Series 13, No. 91. DHHS Pub. No.(PHS) 87-1752. Public Health Service. Washington. U.S. Government Printing Office, May 1987.

Documents

The effect of coordinated, multidisciplinary ambulatory care on service use, charges, quality of care and patient satisfaction in the elderly