For-Profit Enterprise in Health Care (1986)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

For-Profit Enterpnse in Health Care. 1986. National Academy Press, Washington, D.C. Hospim1 Ownership and Be Practice of Medicine: Evidence from He Physicnn's Perspective Robert A. Musacchio, Stephen Zuckerman, Lynn E. Jensen, and Tarry Freshnock The last two decades have witnessed sig- nificant changes in the structure of the hospital industry. While the total number of investor- owned hospitals has remained reasonably sta- bIe since the late 1960s, there has been sig- nificant growth in the number ofinvestor-owned hospitals that are a part of multihospital sys- tems. These hospitals have grown at an av- erage annual rate of 10.3 percent between 1972 and 1983. Further, the concentration of the proprietary sector is quite high: the five largest chains control over 60 percent of all general acute care hospitals in the sector. Concern over the implications of an emerg- ing "medical-industrial complex" has raised a number of questions and sparked considerable debate (Relman, 1980, 1983~. Issues pertain- ing to ethical provider behavior, quality of care, and the comparative economic performance of proprietary and not-for-profit hospitals have recently emerged (Lewin et al., 1981; Pattison and Katz, 1983; Sloan and Vraciu, 1983; arid Sloan arid Becker, 1985~. Ermann and Gabel (1984) have recently published a comprehen- sive assessment of the issues and empirical findings on the subject of multihospital sys- tems. Their review found, among other things, that Mere are no discernible differences be- tween systems and independent hospitals in terms of access, service availability, and qual- ity of care. Thus far, all studies comparing the perfor- mance of systems versus independent hospi- tals and for-profit and not-for-pro~t hospitals Dr. Musacchio is Director, Department of Health Systems Analysis, Center for Health Policy Research; Dr. Zuckerman is Research Associate, Health Policy Center, The Urban Institute, Washington, D.C.; Dr. Jensen is Vice President, Health Services Policy Group; and Dr. Freshnock is Vice President, Lance V. Tarr- ance and Associates, Chicago, Illinois. 385 have focused on the hospital as the unit of analysis. Given the physicians' role as the cen- tral decision maker in the provision of health care services, additional insight can, however, be gained from analyzing physician involve- ment with for-profit hospitals from the phy- sicians' perspective. Until recently the data necessary for such an analysis were unavail- able. Data collected through the American Medical Association's (AMA's) Socioeconomic Monitoring System (SMS) and attitudinal re- search program allow us to examine these is-  sues. This paper describes the data sets used in the analysis, reviews differences in physician characteristics and practice patterns across hospital types, considers the variation in phy- sicians' financial arrangements with hospitals, and examines physicians' attitudes toward for- profit hospitals. Finally, it summarizes our findings and suggests some areas for future DATA Data are drawn from three sources: (1) the 1984 SMS Core Survey; (2) the 1984 AMA Physician Opinion Survey; and (3) the Amer- ican Hospital Association's (AHA's) 1982 AHA AnnualSurveyofHospitals. SMSisa quarterly telephone survey program that collects infor- mation from a random sample of nonfederal patient care physicians stratified by specialty and census division. Data collected include information on physician incomes, practice ex- penses, visits, hours worked, and fees, among other socioeconomic indicators. The core sur- vey is conducted during the second quarter of each year and uses a sample approximately three times as large as those used in other quarters. In 1984, 4,002 physicians were interviewed

386 during the core survey. This survey included a supplemental series of questions on physi- cians' involvement with hospitals. These ques- tions were developed in collaboration with staff of the Institute of Medicine for the purpose of analyzing differences among physicians in for- profit and not-for-profit facilities. Specific areas covered include the number of hospital pnv- ileges each physician has, the physician's fi- nancial arrangement with his/her primary hospital, the physician's evaluation of certain aspects of this hospital, and physician own- ership interest in their primary hospital. This type of information is useful to determine the overall extent of physician involvement with for-profit enterprises and multihospital sys- tems. In addition, SMS also collected information on the physician's primary hospital so that hos- pital data from the 1982 AHA Annual Survey, the most recent one available at the time the research was conducted, could be merged with the SMS file. Merging data from both sources provided us with higher quality information on hospital ownership, organizational struc- ture, utilization, and hospital capacity than would have been available from the physician. While we are aware that some hospitals may have been acquired by for-pro~t organizations between the time of the AHA survey and the SMS survey, we fee] the procedures we fol- lowed gave us the best possible physicianlhos- pital data set we could create. Details on the AHA Annual Survey of Hospitals may be found in Mullner et al. (1983~. For each SMS physician, the unit of analysis in our study, the hospital about which we are collecting information is the one at which the physician provides the most patient care ser- vices. Since the SMS sample is limited to non- federal physicians, only two federal hospitals were picked up. As a result, the federal hos- pital classification has been excluded from our analysis. In addition, long-term hospitals have been excluded for reasons of homogeneity. Findings from the 1984 Physician Opinion Survey are based on telephone interviews with 1,000 randomly selected physicians residing in the United States. The interviews were con- ducted during July 1984. The information pre- sented here is part of a major attitudinal data base which is summanzed in Physician and FOR-PROFIT ENTERPRISE IN HEALTH CARE Public Attitudes on Health Care Issues, avail able Tom the AMA upon request. Table 1 compares AHA data to SMS data with regard to selected hospital characteris- tics. Several differences exist. Specifically, SMS appears to be picking up fewer nongovern- ment for-profit hospitals than the AHA survey. Despite the fact that almost 15 percent of all hospitals are categorized as for-profit, only 8.4 percent of the SMS physicians have their pri-  mary privileges at a hospital of this type. Fur- thermore, hospitals identified through SMS have higher annual admission rates and larger average bed sizes than the hospital population in general. We believe that these differences exist for two primary reasons. First, a large urban/rural difference exists between the two data sets. About 81 percent of the physicians in the SMS data set work primarily in hospitals in urban areas. In comparison, AHA data in- dicate that only 55 percent of the hospitals are located in these areas. Second, and probably more important, is that physicians' primary hospitals the hospitals at which they provide the most patient care services are not rep- resentative of hospitals in general. Due to the urban/rural differences and the characteristics of the physicians' primary hospital, the hos- pitals identified in this analysis do not mirror those in the AHA annual survey. PHYSICIAN CEIARACTERISTICS AND PRACTICE PAI~lERNS An analysis of physician involvement with for-profit hospitals should start by asking the question: Do physicians who have their pri- mary privileges at for-profit hospitals dyer Bom other physicians in any systematic way? In addition, it will be useful to consider if physicians' practice patterns vary according to the type of hospital the physician works in. Given this focus, our analysis is limited to those physicians with admitting or medical staffpriv- ileges at some hospitals. Over 95 percent of physicians in the SMS sample had hospital . . prove eyes. Table 2 shows the specialty distribution of physicians by hospital type. A chi-square test causes us to reject at the 99 percent confidence level the null hypothesis that physicians' spe- cialty and hospital type are independent. Pri

THE PHYSICIAN'S PERSPECTIVE TABLE 1 Comparison of Hospitals in American Health Association (AMA) Annual Survey to Those Represented by Physicians in SMS Sample Hospitals Hospitals Represented Represented by in the AHAa Physicians in Variable Annual Survey the SMSb Sample Hospital type (%) Government nonfederal 29.8 16.2 Nongovernment not-for-profit 55.5 75.4 Nongovernment for-profit 14.7 8.4 Organizational structure (%) Multihospital system 30.0 34.3 Independent hospital 70.0 65.1 Region (%) Northeast 14.8 22.3 Norm Central 28.7 23.0 South 37.5 335 West 18.9 21.2 Location (%) Urban 53.9 80.5 Rural 46.1 19.5 Hospital size Average bed size 168.8 364.0 Admissions per year 5,955 13,563 aAmerican Hospital Association (AHA). b Socioeconomic Monitonng System (SMS) of He American Medical Association (AMA). SOURCE: 1982 AHA Annual Survey of Hospitals and 1984 AMA Socioeco- nomic Monitoring System Core Survey. vale not-for-profit hospitals are staffed by a more specialized group of physicians than other hospitals. Nearly one out of every four phy- sicians whose primary privileges are at a gov- ernment or for-profit hospital is a general or family practitioner (GP/FP). On the other hand, only one out of eight physicians primarily af- filiated with private not-for-profit hospitals is GP/FP. In addition, we final that physicians whose primary hospital is not-for-profit are twice as likely to be in some internal medicine sub- specialty as physicians at for-profit hospitals; 18.5 percent as compared to 9.2 percent. To the extent that specialists treat different types of patients than GP/FPs, these specialty-mix differences may be indicative of variation in hospital case mix. In addition to the growth in for-profit hos- pitals, the expansion of multihospital systems (MHS) represents a major change taking place 387 in the hospital industry. Table 3 indicates that 34.3 percent of physicians have privileges at a hospital that is part of an MHS. There is a significant relationship between being part of an MHS and the ownership status of the phy- sician's hospital. Only 17 percent of physicians primarily affiliated with government hospitals are in an MHS. The MHS structure is most prominent among for-profit hospitals. Sev-  enty-five percent of the physicians with priv- ileges at these hospitals claim MHS affiliation. To explore further differences among phy- sicians in for-profit hospitals and those in other facilities, data on selected physician charac- teristics are presented in Table 4. There are no statistically significant differences among physicians on the basis of experience (years since residency) across the hospital types shown. However, significant differences exist with re- gard to location of the medical school, em

388 FOR-PROFIT ENTERPRISE IN HEALTH CARE TABLE 2 Specialty Distribution of Physicians by Hospital Types (percentage) Hospital Type Specialty Nonfederal Private Government Not-for-profit For-profit Generallfarnily practice 23.3 12.9 22.2 Internal medicine 12.7 18.5 9.2 Surgery 20.0 22.1 24.9 Pediatrics 7.2 S.4 5.5 Obstetrics/gynecology 7.6 7.8 8.5 Radiology 8.1 6.6 7.5 Psychiatry 4.1 6.6 5.5 Anesthesiology 4.6 5.8 6.5 Pathology 4.8 3.5 2.4 Other specialties 7.6 7.8 7.9 NOTE: Chi-square (18 d.f.) = 80.23, which is significant at the 99 percent level of confidence. SOURCE: 1984 AMA Socioeconomic Monitoring System Core Survey. ployment status, and type of practice. Foreign medical school graduates (FMGs) represent a greater proportion of physicians primarily af- filiated with independent for-profit hospitals (31.5 percent) than they do at any other type of facility. For-profit hospitals that are part of an MHS do not have medical staffs that are as heavily dependent on FMGs. Only 20 percent of these physicians received their medical training abroad. This is quite comparable to physicians whose primary privileges are at in- dependent not-for-profits (19.5 percent), in- depen~dent government hospitals (18. 7 percent), and MHS not-for-profits (17.7 percent). The data also show that physicians who are full or part owners of their main practice- self-employed physicians account for 88.2 percent of physicians primarily affiliated with MHS for-profit hospitals. This is a higher pro- portion than we found among physicians at any over types of hospitals. Other types of hos- pitals with above-average shares of self-em- ployed physicians are MHS not-for-profits (81.3 percent) and independent for-profits (80.8 per- cent3. Employee physicians, on He other hand, are most prevalent among government hos- pital staffs, making up 32.3 percent of He MHS group and 26.2 percent of the independent hospital group. These results suggest Hat for- profit hospitals have not attempted to influ TABLE 3 Distribution of Physicians Between Multihospital Systems and Independent Hospitals by Hospital Ownership Status All Nonfederal Private Type Hospitals Government Not-for-profit For-profit Muldhospital system (%) Independent hospital (%) 65.7 83.0 66.5 24.9 NOTE: Chi-square (2 d.f.) = 292.3, which is significant at the 99 percent level of confidence.  34.3 17.0 33.5 75.1 SOURCE: 1984 AMA Socioeconomic Monitoring System Core Survey.

THE PHYSICIAN'S PERSPECTIVE 389 TA:BLE 4 Selected Physician Characteristics by Hospital Ownership Status Percentage of physicians) Multihospital System Independent - Private Private Nonfederal Not-for For- Nonfederal Not-for For Vanable Government profit profit Government profit profit Years since residency 0-5 years 20.8 17.8 18.7 18.4 17.0 15.1 5-10 years 22.9 19.1 19.6 19.2 20.7 13.7 10-20 years 24.0 24.7 30.1 27.8 27.1 30.1 20-30 years 17.7 20.7 20.1 20.9 18.3 19.2 30 or more years 14.6 17.7 11.4 13.7 17.0 21.9 Chi-square (20 d.f.) = 18.5 Location of medical school U.S. or Canada 76.0 82.3 80.0 81.3 80.5 68.5 Other foreign country 24.0 17.7 20.0 18.7 19.5 31.5 Chi-square (5 d.f.) = 9.9* Board certification No 31.3 31.4 38.6 36.2 31.3 37.0 Yes 68.7 68.6 61.4 63.8 68.7 63.0 Chi-square (5 d.f.) = 9.0 Employment status in main medical practice Employee 32.3 18.7 11.8 26.2 ~.1 19.2 Self-employed 67.7 81.3 88.2 73.8 77.9 80.8 Chi-square (5 d.f.) = 29.7*** Type of practice Solo 52.7 46.3 57.4 50.1 51.4 58.8 Non-solo 47.3 53.7 42.6 49.9 48.6 41.2 Chi-square (5 d.£ ) = 12.4** *Significant at the 90 percent level of confidence. **Significant at the 95 percent level of confidence. ***Significant at the 99 percent level of confidence. SOURCE: 1984 AMA Socioeconomic Monitoring System Core Survey. ence physician practice patterns by placing the physician in an employee's role. Other finan- cial mechanisms that could alter practice pat- terns are available to hospitals. Some of these will be discussed in the following section of this paper. For-profit medical staffs also have a signif- icantly higher proportion of solo practitioners Han government or not-for-profit hospitals. Ibe implications of this finding are not clear unless one can argue that practice patterns of group physicians are somehow more in tune with the objectives of not-for-profit hospitals than they are with other hospitals' objectives. We would not be willing to make that argument here. If, however, the division of physicians between hospitals becomes increasingly a function of practice type, this area should be explored fur- ther. At present we do not see this as a major issue. While we found that for-profit MHS medical staffs have the lowest extent of board-cer~fied physicians, this result was not statistically sig- nificant. However, since some view board cer- tification as an indicator of the quality of care delivered at a hospital, we explored this issue  farther (Table 5~. We were particularly con- cerned that the significant specialty composi

390 tion differences might be confounding our analysis of board certification, i.e., there might be significant differences in board certification for certain specialties but not for others. In examining board certification by specialty and hospital ownership status, we found statisti- cally significant variations. In particular, we found that surgeons at for-profit MISS hospi- tals were the least likely group of surgeons to be certified. Since the extent of board certi- fication clearly varies by specialty, general conclusions about the relationship between board certification and hospital type cannot be drawn. Aspects of the relationships that exist be- tween physicians and hospitals may also vary by hospital type. Data on hospital tenure, per- cent of time spent at the primary hospital, the total number of hospital privileges, and the percentage of physicians having solo admitting privileges are displayed in Table 6 as a means for studying these relationships. The data in- dicate that physicians with a primary for-profit affiliation have been working at this hospital for fewer years than physicians at both not- for-profit and government hospitals. We find this result particularly interesting in light of the similarity in physicians' years since resi- dency across hospital types (see Table 41. The lower tenure suggests that the shift in physi- cians' hospital practices to the for-profit sector is a relatively recent phenomenon. This result FOR-PROFIT ENTERPRISE IN HEALTH CARE underscores from the physicians' perspec- tive that the expansion of the for-profit hos- pital sector is fairly recent. If we analyze the data by system and nonsystem status we ob- serve that the only real difference exists be- tween the multifacility for-profit hospital and its independent counterpart. Physicians at a for-profit multifacility hospital have been af- hliated with their hospital the fewest number of years (8.4~. Furthermore, physicians at for-profit hos- pitals have a greater number of hospital priv- ileges than their not-for-profit counterparts. The greater number of privileges may be ne- cessitated by the lower average bed size offor- profit hospitals. Although for-profit facilities have fewer beds on average, physicians at these for-profit facilities spend about the same nllm- ber of hours at their primary facility as phy- sicians at nonprofit hospitals. An alternative way to analyze physician hospital affiliations is to examine the percent of physicians with priv- ileges at only one hospital. Overall, approxi- mately 37 percent of the physicians surveyed indicate that they have solitary admitting priv- ileges. Physicians primarily affiliated with an independent for-profit hospital are least likely to have privileges at only one hospital. At the other extreme, physicians at independent non- federal government hospitals are Me most likely  to have single admitting privileges. As noted, there has been considerable in TA;B~ 5 Percent of Physicians That Are Board Certified by Hospital Ownership Status and Selected Specialty Breakdowns Multihospital System Independent Private Private Nonfederal Not-for- For- Nonfederal Not-for- For Government profit profit Government Profit profit All physicians 68.7 68.6 61.4 63.8 68.7 63.0 Gener~/family specialty b 43.4 40.8 53-9 43.1 b Medical specialty b 67.3 b 65.0 70.3 b Surgical specialty 76.5 79.3 66.7 73.8 76.4 85.0 Othera 75.9 70.1 67.2 63.2 - 70.0 73.9 NOTE: The variation in board certification by specialty and hospital status was found to be statistically significant using a chi-square test. aIncludes psychiatrists, radiologists, anesthesiologists, and pathologists. bInsu£icient number of observations to provide reliable estimates.

THE PHYSICIAN'S PERSPECTIVE TABLE 6 Dimensions of the Physiciar~/Hospital Relationship by Type of Hospital Multihospital System Independent Private All Nonfederal Not-for Hospitals Government Number of years affiliated with primary hospital F test (5 d.£) = 10.6** Percent of hospital hours spent at primary hospital F test (5 d.f.) = 2.84* Number of hospitals at which physician has hospital privileges F test (5 d.f.) = 19.5** Percent of physicians having privileges at only one hospital F test (5 d.f.) = 19.5** 12.3 11.0 12.9 79.8 83.4 77.9 2.1 1.9 2.4 36.7 38.5 26.4 27.3 *Significant vanadon at the 95 percent level of confidence. **Significant vanabon at the 99 percent level of confidence. SOURCE: 1984 AMA Socioeconomic Monitonng System Core Survey. terest in the influence of hospital ownership on physician practice patterns. Table 7 shows that physicians' hours, visits, and hospital dis- charges vary according to the type of hospital at which they have their primary privileges. Physicians at independent nonfederal govern- ment hospitals work on average more total hours, have more patient visits, and discharge more patients from the hospital each week than physicians at the other types of hospitals. Phy- sicians in multifacility for-profit hospitals see the fewest number of patients (22.7) on hos- pital rounds. This relatively low level of uti- lization in the hospital is somewhat offset by the fact that these physicians spend more time and see more patients in their offices. (Despite these differences in utilization, physicians' earnings do not vary significantly by hospital type.) Another issue in the debate regarding hos- pital ownership and performance is whether payer mix in for-profit hospitals is different from the payer mix in other hospitals. Using the hospital as the unit of observation, Sloan and Vraciu (1983) found no significant differ 391 For- Nonfederal profit profit Government Private Not-for- For profit profit  8.4 10.8 12.9 10.2 79.7 82.7 80.1 80.9 2.7 1.8 2.0 2.6 48.3 40.5 20.5 ences in Medicare and Medicaid days as a per- cent of total acute patient days by ownership class. To analyze this issue from the perspective of the physician, we examine whether varia- tion exists in the number of inpatient visits that the physician makes with Medicare pa- tients and the number of hospital discharges by hospital type. The data indicate that there is no significant variation in the weedy num- ber of inpatient visits by hospital type. In ad- dition, while there appears to be variation in the number of Medicare hospital discharges by hospital ownership status, it appears to be consistent with the pattern for total dis- charges. Taken together, these two results suggest that physicians in both for-profit and not-for-profit hospitals may be seeing their Medicare inpatients more frequently than physicians in government hospitals. As such, one might be led to conclude that concerns about the treatment of Medicare patients in for-profit hospitals may be unwarranted. The real distinction appears to be between hospi- tals in the public and private sectors.

392 oo - Ct ._ U) o 4- o c) o ._ ._ - ._ C~ ._ C~ ._ :^ 4- o u) a~ U. a; - Q) cn c: ~s U. C~ .= C~ _ _ - . Ct U) _ C~ oo CC  , C~ ~ o U: C~ 'S ~ C:: z _ ~ CC 0 o o Z C' - ~o ~ CO Le _ C~ Z C¢ ~ C) o o Z C: U. C~ - . V' ~ Ct C~ * * * * * oo _ _ oo CC C~ =4 _ =, ~14 o _ _,,C, ~ . . .. . ~ _ _~ C~ CD ln _c~ ~ 0CD . . .. . .. . o C~o ~ C~C~ _ CO oo C~ C~ _ - = - c~o~ocom~ CD (D OC~ C~ ~CO O C~d4 - U: C~ _~ ~ C~C~ ~ - CD ~ 0e~ ~ c~GO ~ ~O . . .. . .. . .. . 000 C~ U:~ 0 ~C~ C~ C~ oo CO~ C~ _ - C.: - CD~ CD ~O CD -~ ~ CD O 00~ 00 C~ C~ O~ - C~C~ 0O ~C~ ~ _ o, ~ ~ CC CD 0- ~ 0~' C~CO ~ C~ C] O CDCD - C~ . . .. . . ~ 00 0U) oo CO U: C]C~ CD C~ -  oo _ 0~ 0 _ CD ~ 0- ~ 0 C~ ~ C~ C) 3 - ._ ._ 3 3 s~ o 3 =,~ U) C~)·_ § ~ m~C~ ~ ., E~ O ~ ~ E~ O ~ > E~ O _ ~ _0 0 o C~ _ CO C] _ 00 . . .. . CD ~ C~ C~ C~ _ _ ~ ~oo 0 . .. . o ~ _ C~ _ :~ ~o ct ~ - c~ ce _ 3~ _ ~ ~ -; Ct , ~C U, ._ U) _ C) U2 _ ~o .^ ~ C ~ ;` ~ U, ° e, o s" o .- o _ ~ ~z .O  C) ~ o ~8 0 C) - o _ et o . - .= C~ - Ct C~ .= C~ *

THE PHYSICIAN'S PERSPECTIVE FIN^C~ ARRANGEMENTS One area which some analysts feel may be altered by the growth in the for-profit sector of the hospital industry is physicians' financial arrangements with hospitals. Shorten and Evashwick (1981) found that arrangements whereby physicians are compensated by the hospital are more common in for-profit rather than voluntary hospitals. Steinwald (1983), on the other hand, found that hospital ownership had little bearing on the compensation meth- ods of the hospital-based physicians. Based on Steinwald's earlier research (Steinwald and Neuhauser, 1980), he had expected for-profits to "attract physicians by providing relatively lucrative and unencumbered working condi- tions." This hypothesis, however, was not sup- ported by his data. In this study, we examine three specific types of financial arrangements that physicians may have with hospitals. First, we consider the proportion of physicians who received some part of their 1983 income directly from their primary hospital. This may be viewed as the most basic category of physician/hospital con- tract. Second, we measure the extent of lease agreements that physicians have with hospi- tals. Under a lease agreement the physician or his practice compensates the hospital for the use of its facilities or services. Finally, we consider a phenomenon unique to the for-profit sector: the degree to which physicians have an ownership interest in their hospital. Table 8 presents data on the first two ar- rangements to be considered. Column3 shows that 23.5 percent of all physicians in our study either receive some type of direct payment for their hospital or had a lease agreement. This percentage varied significantly by hospital type. However, upon examination of direct nav- ments and lease agreements separately we find that only the percent receiving direct pay- ments showed significant variation. The prob- abiLty of receiving a direct payment was highest for physicians at government hospitals (25.1 percent) and lowest for physicians at for-profit hospitals (11.4 percent). These results are sup- portive of Steinwald's hypothesis regarding "unencumbered working conditions" in for- profit hospitals. In particular, we find that phy- sicians affiliated with hospitals that are a part 393 of for-profit chains are the least likely to re- ceive payments from their hospital. To control for differences in physician char- acteristics, we examine differences in financial arrangements using a linear probability model. Among the control variables we include phy- sician specialty, location of medical school, board certification, employment status, sex, experience, type of practice, and hospital bed size. Holding these variables constant, we find (Table 9) that physicians in for-profit chains,  not-for-profit chains, and independent not-for- profit hospitals are significantly less likely to receive direct payments than physicians in in- dependent government hospitals (the omitted category). This supports the results of our uni- variate analysis. Lease agreements continue to remain unrelated to hospital ownership. Data on the methods by which the hospital's direct payments are made are shown in Table 10. Despite the fact that variation by owner- ship status and organizational structure is not significant, the table provides useful infor- mation on compensation arrangements. The numbers shown do sum to over 100 percent across the rows since each physician may re- ceive payments through a number of arrange- ments. Our results indicate that salary and fee- for-service compensation are the most com- mon methods. Overall, 59.0 percent of those physicians receiving payments were paid some of it in the form of salary, while 37.7 percent received some fee-for-service payments. In this study, we define fee-for-service to include ar- rangements in which the physician receives a fixed percentage of a hospital's charges. Even though there is a lack of statistical sig- nificance regarding compensation and orga- nizational structure, it appears that government and not-for-profit hospitals are different from for-profits with regard to the methods of com- pensation. In particular, the physicians at for- profit chains seem least likely to be involved in a salaried relationship and most likely to receive payments on a fee-for-service basis or as a percent of net or gross department bill- ings. These latter three methods can be viewed as incentive arrangements that are designed to align the financial interests of the physician with those of the hospital. Unfortunately, firm conclusions about the relationship between hospital type and methods of compensation

394 FOR-PROFIT ENTERPRISE IN HEALTH CARE TABLE 8 Physicians' Financial Arrangements with Hospitads by Hospital Ownership Status and Organizational Structure (percentage of physicians) Direct Payments or Lease*** Direct Lease Payments* Agreement** All physicians 18.4 7.1 23.5 Nonfederal government 25.1 7.9 31.1 Multihospital system 24.2 5.2 28.4 Independent hospital 25.3 8.5 31.7 Private not-for-pro~t 17.8 7.1 22.6 Multihospital system 16.5 6.0 21.1 Independent hospital 18.4 7.6 23.3 For-profit 11.4 5.6 16.5 Multihospital system 8.8 5.6 14.2 Independent hospital 19.4 5.5 23.3 *3-way chi-square (2 d.f.) = 26.77; Away chi-square (5 d.f.) = 32.28, both significant at the 99 percent level of confidence. **3-way chi-square (2 d.f.) = 1.55; Away chi-square (5 d.f.) = 4.99, neither significant. ***3-way chi-square (2 d.f.) = 27.39; 6-way chi-square (5 d.f.) = 31.91, both significant at the 99 percent level of confidence. SOURCE: 1984 AMA Socioeconomic Monitoring System Core Survey. cannot be drawn from this study due to the small number of physicians in our sample who have Foliations with for-profit hospitals and receive direct payments. Data not shown in the tables indicate that few physicians with their primary privileges at for-profit hospitals have any ownership inter- est in the hospital. Fewer than one out of ten physicians on for-profit staffs indicated that they were Fill- or part-owners of their hospital. However, we do find that hospital ownership is significantly more prevalent among physi- cians whose privileges are at independent for- profit hospitals. Twenty-two percent of these physicians were hospital owners as compared to only 6 percent of physicians primarily afEl- iated with chain hospitals. In the case of chain hospitals, hospital ownership includes owning stock in the company that controls the hos- pital. PHYSICIAN ATTITUDES In the preceding sections we examined the for-profit issue by analyzing a range of socio economic variables. We found some substan- tive differences in the areas of medical school location, employment status, utilization, and financial arrangements. However, an equally important avenue for analysis is physicians' at- titudes toward for-profit hospitals. If physi- cians, for whatever reason, were to oppose the for-profit concept, then conquer growth in this sector could produce increased tension among medical staffs and hospital administra- tors. In considering these attitudinal results, it is important to remember that fewer than 10 percept ofthe physicians have their primary  privileges at a for-profit hospital. Therefore, many of the opinions regarding for-profit hos- pitals are not likely to be based on firsthand experience. Table 11 presents results from the 1984 AMA Physician Opinion Survey. The exact ques- tions that were asked and the filll range of potential responses are included in the table. According to question 1, only 52 percent of physicians believe that hospitals can be op- erated properly on a for-profit basis. This per

THE PHYSICIAN'S PERSPECTIVE TABLE 9 Determinants of Physicians' Financial Arrangements with Hospitals: A Multivanate Linear Probability Mode] (ordinary least squares estimate: N = 3,023) Dependent Variable - Direct Independent Direct Lease Payment Variablesa Payment Agreement or Lease Physician Characteristics Specialty Internal medicine0.14**0.000080.14** (6.56)(0.01)(5.46) Surgery-0.07**0.02- 0.05* (3.73)(1.12)(2.04) Pediatrics- 0.02- 0.01- 0.02 (0.66)(0.70)(0.69) Obstetncs/gynecology- 0.05*0.0009- 0.05 (1.99)(0 05)(1.53) Radiology0.01- 0.010.01 (0.39)(0.57)(0.34) Psychiatry0.04- 0.010.03 (1.43)(0.35)(0.84) Anesthesiology0.010.010.02 (0. 17)(0~45)(0.66) Pathology0.49**- 0.00090.48** (12. 14)(0.03)(10.32) Other0.11**0.06**0.16** (4.07)(2.71)(5. 11) Years since residency 0-100.050.050.08 (0.76)(0.91)(0.98) 10-200.030.020.04 (0.51)(0.32)(0.44) 20-300.020.010.02 (0.33)(0. 11)(0.24) 30 or more0.001- 0.01- 0.01 (0.02)(0.09)(0.13) Board certification0.0003- 0.0007- 0.004 (0.02)(0.07)(0.24) Foreign medical school0.010.0040.001 graduate(0.75)(0.31)(0.06) Sex (male)- 0.040.03- 0.02 (1.74)(1.43)(0.88) Self-employed0.07**-0.03*0.03 (3.66)(2.22)(1.55) Solo practitioner-0.003- 0.02*- 0.02 (0.22)(2. 16)(1.58) Hospital Characteristics Bed size-0.000020.00001- 0.00002 (1. 13)(0.65)(0~53) (`continued) 395

396 FOR-PROFIT ENTERPRISE IN HEALTH CARE TABLE 9 Continued Dependent Variable Direct Independent Direct Lease Payment Vanablesa Payment Agreement or Lease Ownership and structure Government,- 0.03- 0.01- 0.04 mul~hospital(0.71)(0.25)(0.74) system Not-for-profit,- 0.06**- 0.01- 0.07 independent(3.23)(0.25)(3.11) hospital Not-for-pro~t, muld-- 0.05*- 0.02- 0.06 hospital system(2.34)(1.10)(2.60) For-profit,- 0.020.01- 0.02 independent(0.46)(0.20)(0.49) hospital For-profit,- 0.10**- 0.02- 0.10 mul~hospital(3.45)(1.00)(3.16) system Constant0.080.080.17 (1.11)(1.~)(1.~) it-square0.110.020.09 aAll variables are categorical variables that equal 1 if the physician has the indicated characteristics; O otherwise (t-statistics are shown in parentheses). *Significant at the 95 percent level of confidence. **Significant at the 99 percent level of confidence. cent differs among physicians depending on their experience with the for-profit sector. Among physicians reporting some staff privi- loges at a for-profit hospital, 64 percent favor the for-profit concept. Only 47 percent of the remaining responded affirmatively to this question. More interesting, however, is the fact that 28 percent of the group with for-profit experience oppose the for-profit approach. Unfortunately, data from the Physician Opin- ion Survey could not be disaggregated in a way that allows us to contrast physicians practicing in MHS hospitals to those in independent hos- pitals. Question 2 asks physicians to compare the quality of care in for-profit versus not-for-profit hospitals. The pattern of responses is similar to that for the preceding question. A slight majority of physicians believe quality is un- affected by the profit-nonprofit distinction. However, views toward quality differences are also affected by actual experience in for-profit hospitals. Twenty-four percent of physicians who have some involvement in for-profit hos- pitals say quality of care is better in the not- for-profit sector. Part of the reason that phy- sicians may view quality as being lower in the for-profit sector is that only 9 percent of them believe that for-profit hospitals afford them a greater degree of clinical discretion (question 31. Changes in the health care sector are caus- ing new ethical issues to emerge. A substantial majority of physicians believe that referring patients to facilities in which they have an own- ership share is a conflict of interest. Physicians  who have some experience in for-profit hos- pitals are less likely to indicate conflict of in- terest. However, even among this group, a majority of 59 percent agree with the conflict- of-interest position. This potential for conflict of interest may be the major reason why few physicians have ownership interests in hos- pitals. As the for-profit hospital sector continues to grow, concerns intensify about graduate med- ical education and its evolving role in an en

THE PHYSICIAN'S PERSPECTIVE vironment increasingly affected by profit considerations. In question 5, physicians were polled concerning their views toward estab- lishing residencies in profit-oriented hospitals. As indicated, a plurality of physicians believe that the quality of graduate medical education would be negatively affected by locating it in a for-profit hospital. Actual experience in for- profit hospitals malces no difference in the phy- sicians' opinion on graduate medical educa- tion. Physicians were then asked about their pref- erence for practice in profit versus nonprofit hospitals. The most prevalent response among physicians interviewed is that no preference between for-profit and not-for-profit hospitals exists, but of those with a preference the not- for-profit hospital was the predominant choice. This seems to reflect the distribution of phy- sicians across hospital types when viewed from the standpoint of their primary affiliation (Table 11. In an effort to evaluate views on the quality of hospitalcare delivered, the SMS survey asked physicians to compare their primary hospital 397 to other hospitals that they may be familiar with, along certain specific dimensions. These results are displayed in Table 12. We find that physicians primarily admitting to government MHS hospitals are least likely to view their facility as better than other facilities. In fact, almost half of these physicians see their hos- pitals as worse than other other hospitals with respect to technical resources and equipment offered. Among physicians admitting to private- sector hospitals, opinions are mixed as to whether for-profits or not-for-profits are bet- ter. Few physicians at any of the private hos- pitals would rate their hospital as worse than other hospitals. In terms of nursing support, the for-profits get a higher rating than not-for- profits among physicians at independent hos- pitals, but a lower rating among the MHS group. We observe a similar mixed pattern in terms of physicians' evaluation of the level of patient satisfaction. In two areas, physician opinions on He for-profit/not-for-profit comparison are clearer, but are contradictory in their evalu- ation of hospitals. Physicians at for-profit hos TABLE 10 Methods of Compensation by Hospital Ownership Status and Organizational Structure (percentage of physicians) . . . . ' ~0 ~ Share of Net Fee for DepartmentShare of Status/Structure Salary Servicer RevenueGross Other . . All physiciansb 59.0 37.7 8.012.1 24.6  Nonfederal government 60.4 35.5 8.812.6 24.8 Multihospital system 78.3 31.8 9.19.1 36.4 Independent hospital 56.5 36.2 8.?13.3 - .6 Private, not-for-profit 59.5 37.5 7.2ll.S 24.0 Multihospital system 50.4 43.4 4.510.4 25.0 Independent hospital 63.5 35.0 8.312.1 23.6 For-profit 48.5 50.0 15.618.2 31.3 Multihospital system 36.8 57.9 11.126.3 27.8 [~.~ade hospital ~.3 ~.5 21.4 7.1 ~.7 NOTE: In this table, none of the variation by hospital ownership status and organizational structure is statistically significant based on chi-square tests. aIncludes arrangements involving the physician receiving a percentage of hospital charges. bThose receiving part of their 1983 income directly from hospital.

398 FOR-PROFIT ENTERPRISE IN lIEALTH CARE TABLE 11 Physician Attitudes on Issues Related to Hospital Ownership by Extent of Involvement with For-profit Hospitals (percentage) Question All Response Physicians Involvement with For-profit Hospitals Some None 1. As you may know, there has been a rapid growth of"for-profit" multihospital systems in recent years. Do you favor or oppose the view that hospitals can be operated properly as profit-making organizations? 2. In general, do you believe that not-for- propt hospitals provide better, worse, or about the same quality of care as for-propt hospitals? 3. In your opinion, do physicians have greater, less, or the same amount of clinical discretion at a for-profit hospital as they do at a not-for-profit hospital? 4. Do you believe it is a conflict of interest for physicians to refer patients to a hospital or over health care facility in which they have an ownership interest? 5. Do you believe that graduate medical training in for-profit hospitals would be better, worse, or the same as that in not-for-profit hospitals? 6. Assuming Me medical staff privileges, compensation arrangements, and work conditions are the sane, would you rather practice at a for-profit hospital, a not-for-profit hospital, or would you have no preference? Favor Oppose Unsure Better Worse Same Unsure Greater Less Same Unsure Yes No Unsure Better Worse Same Unsure For-profit Not-for-profit No preference 52 37 11  32 5 50 13 9 28 37 27 ·65 28 7 5 46 34 16 7 64 47 28 41 8 12 24 8 59 9 11 27 50 12 59 35 7 44 37 13 13 42 31 51 57 46 14 28 32 32 68 26 6 4 47 34 16 48 47 SOURCE: 1984 AMA Physician Opinion Survey. pitals are more satisfied with the responsiveness of the hospital to their professional needs, but less satisfied with the available technical re- sources and equipment than their not-for-profit counterparts. On the basis of these results, no clear preferences are apparent among private- sector physicians.  SUMMARY AND DISCUSSION The purpose of this paper has been to com- pare characteristics of physicians at for-pro~t hospitals to those of other physicians. In ad- dition, we explored physician attitudes toward aspects of the growing for-pro~t hospital sec tor. At this point in time, physicians have lim- ited experience with for-profit hospitals. Even so, the attitudinal data (Table 11) show that half of all physicians believe hospitals can be operated properly as profit-maldng organiza- tions. However, data on specific areas of hos- pital operation (Table 12) produced mixed results with regard to preferences for partic- ular hospital types. While we realize that our own data could suggest that physicians are not satisfied with for-pro~t hospitals, a more ap- propriate interpretation would be that physi- cians are clearly aware of and concerned about potential issues that could arise. A less ten- tative conclusion is unwarranted until more

399 ct c) - :^ - c~ · - v, o ;^ ct ~ - o ~ - As ·s~ am .~ o o ~ - c~ en ·c) · - Ed - c c)  =5 £ V' ;` Cal - ._ Cal o ._ - o ° o ._ o Z c C~ C) c, £ =5 C (V J c~ Q) O O Z C: - O OJ :> _ Cc: Z _~ eC C ~ £ C.) J O O Z C' C) C c~ ~; _ 0, _ ~ c~ oo - ~4 CD C~ ~ . . . . . . . . . . . Q _ ~ ~ ~ ~ CJ ~ ~ CD C~ ~ C~ ~ C~ C~ ~ U) 00 ~ CO ~ ~ ~ ~ U: CD . . . . . . . . . . . . CD O C~ C~ ~ ~ ~ ~ ~ O o C~ {D c~  O ~ 0, ~ ~ 00 ~ (D o c~ ~ c~ ~ - u: ~ cO ~i u) ~4 ~ - ~ "4 ~ ~ c~ - c~ ~ - c~ oo o ~ c;, ~ u, o u, co ~ O c~ r~ _ C~ "4 C~ e~ ~4 ~r (D C~ ~ ~ ~ ~ C~ O C~ O0 ~ 00 ~ 00 0 ~ ~ ~ c~ 0 co _ _ ~ a' c~ _ o c] ~ ~ - cO ~ ~ c~ ) ~ - c~ ~ c~ oO oD =4 c~ - cO - ~ t- ~ ~' ~ 3_ (J,) 5 ~ ~ ~ U) CD ~ ~ C) C) ~ ~ ~ C~ ,,, ~ ~o ~ £ 3 ~ £ 3 ~ ~ 3 .8 ,, , ,°. ;; ~ e e _ ~Z =; ~ .~ o U, CJ - . - o~ Ct 5: C) c) _ ._ ~5 C~ t4 3 C~ _ .~ c) "S _ - C'3 - C~ ._  ~S O L _ cn _ J ._ O o =: ~ ~2 ~;` '" t0 ._ ~_ ~O 5: ~ _ C ~O ce ~C) ~._ cO' 0~ o o ~ c., O _y ._ c ~O 5- ~ C) 00 3 ~ U) . . _ ~ 'C) ~ ._ ~ ~ O C U)

400 physicians have firsthand contact with these facilities. To summarize, examining the physician characteristic data, we identified a number of important differences between for-profit med- ical staffs and those at other hospitals. First, there are relatively more GP/FPs and rela- tively fewer internists on the staffs offor-pro~t hospitals. This may reflect differences in hos- pital case mix. Second, the typical physician on the staffofafor-profit MHS hospital is more likely to be self-employed than other physi- cians. Third, this same physician is least likely to receive any direct monetary compensation from his/her hospital. Taken together, these last two findings suggest that the financial link between the hospital and the physician may be weakest at for-profit MHS facilities, a result that may be counter to some views of the for- profit sector. Finally, physicians primarily affiliated with for-profit hospitals discharge the fewest pa- tients from the hospital in an average week. This, combined with their higher average rate of office visits, suggests that their practices may be somewhat less involved with the hos- pital than are those of other physicians. The lower level of hospital involvement is quite likely due to the specialty composition ofthese physicians and the case mix of their patients. Our results show that, along several dimen- sions, medical stabs at for-profit hospitals dif- fer from those at other facilities. It appears that the specialty composition differences may be the driving force behind our findings. The higher proportion of GP/FPs relative to in- ternists at for-profit hospitals could result in less-hospital-oriented practices and the larger proportion of self-employed physicians that we observe. For-profit hospitals may be treating comparable types of patients with less spe- cialized physicians. On the other hand, by vir- tue of their specialty differences, for-profit medical stabs may be admitting a different mix of cases to their hospitals. Our data do not allow us- to choose between these competing hypotheses. The relationship between the ob- served specialty differences, case mix, and quality of care is an area that should receive Farther investigation. This study was not designed to be an ex- haustive analysis of the for-pro~t multihospital FOR-PROFIT ENTERPRISE IN HEaLTH CARE system issue. We feel a number of other fruit- ful areas for future research remain open with regard to the physician. One important area that might help us to better understand phy- sician/hospital relations in general is: What fac- tors cause a physician to seek privileges at a particular type of hospital? For example, do physicians affiliate with a for-profit hospital or multihospital system because it is the only one located near them, or are there specific aspects  of the hospital environment that make for-profit hospitals more desirable to some physicians? Likewise, do for-profit hospitals grant hospitaV physician privileges to different types of phy- sicians? A second area of research that seems appropriate in light of our findings with regard to financial arrangements is: As for-profit hos- pitals grow, will new types of financial arrange- ments between physicians and hospitals develop? Physiciar~lhospital joint ventures and physician ownership of stock in hospital man- agement companies are two such develop- ments that have the potential to alter practice patterns and, therefore, should be studied ~r- ther. Finally, will Medicare's prospective pric- ing system, other regulatory devices, or compe- tition change the need for a financial tie be- tween hospitals and physicians. Throughout most of this paper we have at- tempted to array the data so as to allow us to analyze differences among government, not- for-profit, and for-profit hospitals. On the basis of Table 7, we observe that physicians prac- ticing primarily in government hospitals ex- hibit utilization patterns that are no more similar to the not-for-profit group than the for-profit group is to the not-for-profit group. In partic- ular, physicians who practice at government hospitals have a higher rate of Medicare dis- charges. This higher rate may be indicative of a shift in the treatment of the aged towards the public sector. While these results are by no means conclusive, we feel that they suggest a need for farther research on public/private differentials. ACKNOWLEDGMENTS The views expressed in this paper are not necessarily those of the American Medical As- sociation or the Urban Institute and its spon- sors. The authors would like to thank Steven

THE PHYSICIAN'S PERSPECTIVE D. Culler, lanes Wilier, Bradford Gray, and two anonymous reviewers for their helped comments and able assistance on this research project. REFERENCES Ermann, D., and J. Gabel (1984) Multihospital sys- tems: Issues and empirical findings. Health Affairs 3~1~:50-64. Lewin, L. S., A Derzon, and R. Margulies (1981) Investor-owned and nonprofits differ in economic per- formance. Hospitals 55:52~8. Morrisey, M. (1984) The composition of hospital medical staffs. Health Care Management Review (Sum- mer):ll-20. Mullner, Ross M., Calvin Byre, and Cleve L. Kil- lingsworth (1983) An inventory of U. S. health care data bases. Review of Public Data Use 11~21:85-188. Pattison, R. V., and H. M. Katz (1983) Investor- owned and not-for-profit hospitals: A comparison based on California data. New England Journal of Medicine 309:347-353. 401 Relman, A. S. (1980) The new medical-industrial complex. New England Journal of Medicine 303:963- 970. Relman, A. S. (1983) Investor-owned hospitals and health care costs. New England Journal of Mediane 309:370-372. Rosett, R. M. (1974) Proprietary hospitals in the United States. Pp. 57-6:5 in The Economics of Health and Med*al Care, M. Perlman, ed. New York: John Wiley and Sons. Shortell, S. M., and C. Evashwick (1981) The struc- tural configuration of U.S. hospital medical staffs. Med- *al Care 19:419 430. Sloan, F., and E. Becker (1985) Hospital ownership and performance. Economic Inquiry. Vol 11. Sloan, F., and R. Vraciu (1983) Investor-owned and not-for-profit hospitals: Addressing some issues. Health Affairs 2~1):22;-37. Steinwald, B. (1983) Compensation of hospital-based physicians. Health Services Research 18~1~:17 43. Steinwald, B., and D. Neuhauser (1980) The role of proprietary hospitals. Law and Contemporary Prob- lems 35:817-838.