Secondary Logo

Journal Logo


The Effect of a Teaching Hospital's Financial Crisis and Reorganization on a Group of Residents

Kiely, Sharon C. MD, MPM; Russo, Kelly Sipe MD, MPH; Orav, E. John PhD; McMahon, Kelly MD; Shannon, Richard P. MD; O'Donnell, Walter J. MD

Author Information
  • Free


Financial adversity has become commonplace in the American health care system as changes in reimbursement and other issues force market change. The medical1 and health policy2 literatures have reported these events and analyzed their implications for academic medical centers, physician practice management organizations,3 and patients.4 However, the effects of system reorganization and financial adversity on residents have received little attention, despite graduate medical education's influence on a hospital's reputation, research programs, and recruitment of medical staff. Instead, the literature has focused on the effects of market and financial issues on working conditions,5 decisions to subspecialize,6 and employment-seeking experiences.7 Teaching hospitals have recognized the importance of financial issues relative to graduate medical education in general, with a focus on reimbursement and payment issues8 and curriculum change.9 However, to our knowledge, there is to date no study of the effects of a financial crisis on the perceptions of residents themselves. Improved understanding of the perspectives of residents in a financially insecure hospital could provide valuable information to enable residency program directors and other leaders to prepare for program implications and support residents during periods of transition and financial crisis.


The Allegheny Health, Education and Research Foundation (AHERF) was a statewide, integrated delivery system in Pennsylvania, and the parent corporation of Allegheny General Hospital (AGH) in Pittsburgh. In June 1998, the AHERF filed for Chapter 11 bankruptcy protection. At the time of the filing, AGH was a 670-bed teaching hospital and the western campus of MCP—Hahnemann School of Medicine in Philadelphia. AGH and its three affiliates were not among the subsidiaries of the integrated delivery system that joined in the bankruptcy filing. The parties to the filing were restricted to the eastern Pennsylvania (Philadelphia) entities, which included MCP—Hahnemann School of Medicine, and 11 affiliated hospitals and the statewide faculty practice plan and primary care network.10 Nonetheless, the finances of AGH were substantially compromised by the system's collapse. To stabilize the system, AGH and its affiliates merged with The Western Pennsylvania Hospital Health System in August 2000.

The Bankruptcy's Effect on the Hospital

Many changes occurred at AGH because of the bankruptcy of AHERF. Initially, there was enormous financial and psychological tension among employees and staff, as well as among the volunteer and full-time physicians. Abrupt reorganization of the administration, media reports of criminal charges, and downsizing of the primary care network called into question the future of the hospital. During this period, a competing hospital system commissioned a report, published before the merger, contending that the merger would force bankruptcies at other hospitals. If this report had been so influential as to prevent any proposed merger, AGH could have ceased to exist. An interim management agreement and the ultimate merger with The Western Pennsylvania Hospital Health System allayed the financial uncertainty. However, the psychological effects of the upheaval were felt for some time. Attending physicians, faculty, and other personnel had begun leaving their positions at AGH to seek stability in their careers. This was aggravated by local institutions' aggressive recruitment of essential faculty and staff physicians. At the time of our study, 21 full-time and 12 private physicians had left the department of medicine. Programmatic changes due to the physicians' departures included the closure of the medical oncology fellowship. At the time we administered the questionnaire in our study, the program director, three division chiefs (renal, cardiology, oncology), the vice-chair of clinical affairs, and the vice-chair of research had resigned. Finally, students and faculty were uncertain about the future of the bankrupt MCP—Hahnemann School of Medicine. These particular concerns lessened later in 1998, when Drexel University in Philadelphia assumed operations with a two-year agreement that included maintaining enrollment levels and AGH as a training site.

The Bankruptcy's Effect on the Graduate Medical Education Environment

AGH has a long history in graduate medical education and was initially affiliated with the University of Pittsburgh and subsequently with MCP—Hahnemann School of Medicine. At the time of the financial reorganization, there were 26 residency and fellowship programs with 270 residency positions. These programs were highly competitive in recruiting new residents and faculty, and the financial situation immediately affected the residency programs. For example, at the time of the announcement of the bankruptcy, several non-medicine residents chose not to come to AGH, and one neurosurgery resident transferred. Some general benefits were eliminated, including an annual graduation party and small research grants. Residents' travel and dues allowances were cut 50% to $1,500 per resident. The loss of benefits was mitigated because some individual residency programs continued to support these activities. Finally, recruitment was negatively affected in some AGH residency programs.

In general, by 1999, the merger had been successful, resulting in a change in Standard & Poor's outlook on the system from “developing” to “stable.” A number of factors influenced the assessment. These included the overall stability of the system's components, the ability to recruit excellent clinicians, financial results consistent with those forecast at the time of the bond issue, successful strategies that addressed workforce vacancies, plans to rebuild or expand important programs, and success in resolving AHERF-related issues. However, the rebuilding of the full-time faculty and the restoration of academic support lagged.

It was clear to the leadership of the internal medicine program that individual residents had been adversely affected by the disruptions. Many residents talked with faculty about their concerns regarding the reorganization. In the course of this general survey of residents' satisfaction, we sought to learn more about the experiences of internal medical (IM) residents during AGH's reorganization and the effects of these events on the perspectives of the residents.


Participants and Instrument

During a six-week period in the spring of 2000, as the financial issues of the bankruptcy and merger resolved, we asked residents in the three IM subprograms (categorical, primary care, and emergency medicine—internal medicine [EM/IM]) to complete a confidential, voluntary questionnaire. (Our study questionnaire is available on request.) Since that initial survey, all IM residents have been surveyed annually as part of the quality improvement effort of our ambulatory program. In 2000, we included the questions that were the focus of our study because we felt the information that we would obtain would be crucial to maintain residents' satisfaction and our program's effectiveness. All 75 IM residents at AGH were offered the opportunity to participate. We informed them that the information would be used to improve the academic program at AGH. Questionnaires were distributed to residents in the ambulatory clinic and in the Department of Medicine's offices. As responding to the questionnaire was voluntary, doing so implied consent to participate in our study. Residents who chose to participate returned their questionnaires to an envelope in the department of medicine's offices, marked their participation on a roster, and indicated their choices of incentives: a meal ticket or three ambulatory medical education credits. Thus, we had a record of residents' participation that could not be connected to the content of their questionnaires. Three residents on maternity leave received and returned the questionnaire by mail. The information obtained was not recorded or reported in a manner that would have allowed for identifying participants and could not reasonably have placed the participants at risk of damage to their reputations or employability.

Two residents and two attending physicians completed the questionnaire during development and pilot testing. The questionnaire took less than 20 minutes to complete and included questions on demographic information, inpatient and outpatient medical education, and AHERF/AGH financial issues. Demographic questions included postgraduate year, program type, age, sex, home town, medical school, financial debt, service obligation, plans for fellowship, and job choices. The questionnaire asked about factors that had influenced the participants' choice of AGH's residency program and whether the AHERF/ AGH financial issues had influenced the participants' residency experiences. We further explored the effects of the crisis on their attitudes and concerns. The questionnaire also included sections on general inpatient and outpatient medical education that are not discussed in this report.

Data Handling and Statistical Analysis

We entered the questionnaire results into a standard database software. Each response was hand-entered by a single investigator. A team of two other investigators independently confirmed or revised each entry. A participant's failure to follow directions and omissions were coded as blank. The amounts of missing data varied by question and study results were always based on all available responses. All debatable responses and disagreements in data entry were reviewed and adjudicated by consensus. Due to the small number of postgraduate year 4 (PGY-4) residents, this group was combined with the PGY-3 residents into a single cohort.

Statistical Analyses

In this report, we give descriptive results from the questionnaire as numbers and their corresponding percentages. Responses are compared among residents, divided according to program year, by Fisher's exact test for binary responses and by the Kruskal—Wallis test for ordinal responses. For issues on which residents were asked to rate the influence of the financial situation as either “unfavorable,” “unchanged,” or “favorable,” we tested for a change in perception by comparing the number of residents who responded “unfavorable” with the number who responded “favorable” using a binomial calculation with a null hypothesis that there was a 50% chance of responding in either direction. The same test was used to compare the number of residents with “increased” concerns with those with “reduced” concerns about a number of personal and professional issues.

We identified before analysis five questions that we considered study outcomes: (1) Have you considered transferring?; (2) Would you recommend AGH?; (3) Has your opinion about large health care systems become unfavorable?; (4) Has your opinion about medicine as a career become unfavorable?; and (5) Do you have concerns about fellowship opportunities? The relationship between each pair of outcomes was examined by a 2-by-2 table and Fisher's exact test. Then, for each of these binary outcomes, we constructed a logistic regression model based on three sets of predictors: basic demographics, reasons for selecting AGH for residency, and influences of financial issues on various aspects of the residency experience. We used forward selection to build the best possible model within each set of predictors. The significant predictors from each set were then combined into one final model, again by forward selection.

We considered p values less than .05 to be significant, and all tests were two-sided.



Figure 1 depicts the timeline for the events in the financial crisis at the Allegheny Health, Education and Research Foundation and Allegheny General Hospital, including match days and start times for each class of residents in our study. The PGY-2,-3, and -4 classes matched before the bankruptcy was declared. The PGY-2 class started residency shortly after the June 1998 announcement. The PGY-1 class matched at AGH nine months after the announcement.

Figure 1:
Timeline of events in financial crisis at Allegheny Health, Education and Research Foundation and Allegheny General Hospital, Pittsburgh, Pennsylvania, including match days and start times for each class of residents in this study.

Respondents and Non-respondents

A total of 71 residents (95%) completed the questionnaire. The median age of respondents was 30 (range 24–46 years), and 29 (41%) were women. One PGY-1 resident, one PGY-2 resident, and two PGY-5 (ER/IM) residents did not respond. Forty-seven respondents (66%) reported home towns other than Pittsburgh; 22 (31%) reported Pittsburgh and two (3%) reported Philadelphia. A total of 52 respondents (73%) had graduated from medical schools other than the University of Pittsburgh School of Medicine (UPSM), while nine (13%) had graduated from UPSM and ten (14%) had graduated from MCP—Hahnemann School of Medicine. Almost half of all respondents (n = 33; 47%) planned to pursue a fellowship.

Factors Influencing Residents' Choice of AGH for Residency

Respondents cited the clinical and academic programs as the most important factors in the choice of AGH for residency (see Table 1). Little difference was seen between the PGY classes, except that location appeared to be least important to the PGY-1 class (4; 18%), compared with the PGY-2 class (11; 46%), and the classes for PGY 3- and -4 (14; 56%) (chi-square p = .026). Facilities were mentioned as influencing choice by the PGY-2 cohort only (chi-square p = .016).

Table 1:
Factors That Influenced 71 Internal Medicine Residents to Choose Residencies at Allegheny General Hospital (AGH) in Pittsburgh, Pennsylvania, by Postgraduate Year (PGY), 2000*

Influence of the Financial Crisis on Residents

Over all, 55 (77%) of our respondents stated that the financial crisis had had an influence on their residency experiences. We asked the residents “In what ways have the financial issues faced by AGH influenced your residency experience?” They were asked to select a single answer on a five-point Likert scale (1 = none, 2 = minor, 3 = some, 4 = moderate, and 5 = great). More than 75% of respondents cited more than a minor influence on their residency experiences due to changes in attending staff, morale, AGH's reputation in Pittsburgh, and nurse staffing shortages brought about by the financial issues (see Table 2). Differences between PGY groups were statistically significant regarding morale (p = .023), lectures/grand rounds (p = .015), and outpatient experiences (p = .012).

Table 2:
Percentages of 71 Internal Medicine Residents at Allegheny General Hospital (AGH) Who Chose Factors in Their Residencies That Were Influenced by AGH's Financial Crisis, by Postgraduate Year (PGY), Pittsburgh, Pennsylvania, 2000*

A total of 40 respondents (56%) reported that the financial issues had affected them personally; of these, 30 (75%) characterized that influence as negative. More than 80% of the respondents (p < .001) reported increased concern about the future of the residency program and the hospital. Respondents' concerns about opinions of family members were not significant. Financial issues made residents significantly more concerned in a broader sense, including concerns about the opinions of colleagues (p < .001), job opportunities (p < .001), and the financial status of a health care system before employment (p < .001).

Outcome Measures

The five outcome measures were statistically independent (p > .05, Fisher's exact test, comparing responses between the elements of each pair of questions). The strongest relationships were: residents who considered transferring were also the most likely not to recommend AGH (p = .072), the most likely to have unfavorable opinions about medicine as a career recommendation (p = .072), and the most likely to have increased concern about fellowship opportunities (p = .096).

Consideration of transfer. Eighteen respondents (25%) had considered a transfer. Differences between PGY groups were statistically significant (chi-square, p = .014), with PGY-3 and -4 combined being less likely (2; 8%) to consider transferring than either PGY-1 (6; 27%) or PGY-2 (10, 42%) residents. Among the issues influencing residency experiences were five univariate predictors, including nurse staffing, faculty participation in education, morale, educational quality, and AGH's reputation outside Pittsburgh. In multiple logistic regression, three predictors emerged: (1) PGY-3 and -4 residents were less likely than PGY-1 or PGY-2 residents to consider a transfer; (2) respondents who had chosen AGH for the clinical program were less likely to consider transfer; and (3) respondents who had chosen AGH for educational quality were more likely to consider transfer.

Current program recommendation. The majority of those responding, 44 of 59 (75%) said they would recommend AGH to a medical student for postgraduate training. The difference regarding willingness to recommend the program was statistically insignificant (p = .40) according to respondents by PGY group: PGY-1 85%, PGY-2 71%, PGY-3 and -4 67%. Demographics and factors that influenced choice of AGH for residency were not predictive of reluctance to recommend AGH. Among the issues influencing residency experiences were nine univariate predictors: inpatient experience, hospital operations, nursing staff, faculty participation, morale, educational quality, topic mentioned by applicants, and AGH's reputation outside Pittsburgh. In multiple logistic regression, only educational quality emerged as a significant predictor. Residents who reported a greater influence of the financial issues on educational quality were less likely to recommend AGH.

Concern about fellowship opportunities. The concerns about fellowship opportunities of residents (53%) increased significantly (p ≤ .001) because of the financial issues. Primary care and older residents were less concerned than their counterparts. More concerned about fellowship opportunities were PGY-2 residents (70%) compared with PGY-1 (48%) and PGY-3 and -4 (44%), although this difference was not statistically significant. Univariate predictors for a greater concern about fellowship opportunities included noting an effect on the hospital's operations, morale, and the hospital's reputation outside Pittsburgh. In a multiple logistic regression model there were two independent predictors of concern about fellowship opportunities: AGH's reputation outside Pittsburgh and the primary care demographic. Residents who reported the most influence on their residency experiences regarding AGH's reputation outside Pittsburgh were the most likely to have increased concerns about fellowship opportunities. As expected, primary care residents were the least likely to reflect this concern.

Opinions of large health care systems. Residents' opinions about large health care systems became more unfavorable because of the financial crisis (p < .001) (see Table 3). There were no demographic predictors regarding changes in residents' opinions of large health care systems. All of the residency groups, PGY-1 (55%); PGY-2 (65%), and PGY 3- and -4 (56%), had essentially equally unfavorable opinions. There were three univariate predictors (nurse staffing, morale, educational quality) of this unfavorable opinion. In the multiple logistic regression models, those who reported the most effect on nurse staffing shortages were more likely to report unfavorable opinions about large health care systems.

Table 3:
Influences of Allegheny General Hospital's (AGH) Financial Crisis on the Opinions and Attitudes of AGH Internal Medicine Residents Relative to Health Care System and Personal Experience Issues, Pittsburgh, Pennsylvania, 2000*

Opinion of medicine as a career recommendation. Residents' opinions about medicine as a career recommendation also became more unfavorable because of the financial crisis (p < .001) (see Table 3). Residents who were graduates of either the University of Pittsburgh School of Medicine or MCP—Hahnemann School of Medicine and residents who had chosen AGH because they had had medical school experience at AGH noted an unfavorable opinion of medicine as a career recommendation more frequently than did others. Again, the difference among responses by PGY class was not significant. Residents' perception of an unfavorable influence on hospital operations, as a result of the crisis, was the only univariate predictor for an unfavorable recommendation of medicine as a career. In the multiple logistic regression model, respondents who had graduated from the University of Pittsburgh School of Medicine or MCP—Hahnemann School of Medicine and who reported that hospital operations were greatly affected had more unfavorable opinions than did those whose opinions about careers in medicine grew more favorable or were unchanged.


We undertook this study to understand the effects of a financial crisis on our residents. We recognize that the basis for experiences, attitudes, and opinions is multifactorial and would not be accounted for completely by this study. Therefore, we considered several potential outcomes to see whether this common experience resulted in similar responses. We hypothesized that these outcomes might be influenced by predictors such as residents' demographics, their reasons for selecting AGH for residency, and the effect of the financial issues on the residency experience. Our analysis demonstrated that residents' responses to the financial crisis and its consequences were indeed complex, with only a few predictors identified in each of the five outcome measures. There was no single predictor of any of the outcome measures by multiple logistic regression.

Effect on IM Residents

Our survey of IM residents revealed that the financial changes at AGH had a substantial effect on their experiences and opinions. It might appear that no significant disruption occurred, since 29 (41%) of the 71 respondents did not report a personal effect of the crisis. Only one resident transferred (for family reasons), although 18 (25%) considered transferring. Three fourths of the respondents noted they would still recommend AGH for postgraduate training. From a patient care standpoint, one third or less of the residents noted that the crisis had influenced the volume of patients or hospital operations, although a majority had noticed an influence on nurse staffing.

On the other hand, our study showed that more than three fourths of the respondents reported that the crisis had influenced their residencies. More than half of each PGY cohort had observed more than a minor effect of the crisis on teaching faculty, morale, institutional reputation, and nurse staffing.

Demographic Predictors of Effect

More of the residents we surveyed were women (29; 41%) and primary care residents (13; 18%) than would have been the averages for residents in the United States in 1992–93, the most recent years for which we obtained data.11 In our analysis, demographics and year of training were generally not strong predictors of experiencing an effect. The exceptions were that PGY-1 residents were less likely to report an effect, perhaps due to their recent arrivals; PGY-2 residents, in the midst of deliberating about a subspecialty, were significantly more concerned about the effect of the crisis on their fellowship opportunities, although this was not borne out in multivariate analysis; and PGY-3 and -4 residents were less likely to contemplate a transfer, perhaps due to their proximity to graduation.

We found that these residents tended to have chosen AGH for essentially the same reasons, regardless of when they had matched at AGH relative to the financial crisis, with two exceptions. Location appeared more important to PGY-3 and -4 residents, possibly reflecting the increased number of Pittsburgh-based medical school graduates in this group. The residents who had chosen AGH for its clinical programs were less likely to consider transfer, perhaps reflecting that clinical volume suffered little during the crisis. However, residents who had chosen AGH for educational quality were both more likely to consider transfer and less likely to recommend AGH to a medical student.

The percentage of residents in our study planning for fellowship training (47%) was similar to percentages of residents in previously published studies7,12 and close to the U.S. goal of 50%. Those most aware of the financial crisis' effect on the hospital's national reputation were the most concerned about their fellowship opportunities, a reasonable concern for residents entering the national competition for a fellowship position. Not surprisingly, primary care residents were the least concerned about the effect of the crisis on fellowship prospects.

Effect on Residency Experience

The financial issues influenced qualitative more than operational factors related to the residency experience. Qualitative factors (changes in staff, morale, AGH's reputation in Pittsburgh and nationally, nurse staffing, attending physicians' and faculty member's participation, educational quality, and the topic being mentioned by residency applicants) were cited by more than half of the residents as having had “some” to “great” influence on their experiences. Specifically, changes in attending staff was selected most frequently and consistently by all groups of residents as having influenced their experiences. This perception accurately reflected reality with respect to the exceptional turnover in both full-time and voluntary program faculty during the reorganization.

Effect on Residents' Perspectives

We found that the crisis had significantly and paradoxically altered the views of the residents regarding large health care systems. Residents' opinions became less favorable toward large health care systems because of their experiences with the collapse of their own system. Interestingly, we found that the residents who had no direct experience with for-profit health care systems in this crisis or in the region, nonetheless asserted that they now looked upon for-profit organizations more unfavorably. On a more optimistic note, nearly half looked more favorably on physicians' involvement in the administration of health care systems, although we cannot be sure that these residents would have been inclined to take such leadership positions. While the disenchantment with large health care systems (and potential employers) is concerning, we found that the residents felt this crisis very personally. They had increased concerns about the opinions of colleagues, and about fellowship and job opportunities. Most were more concerned about the future of the program, and indeed of the hospital, because of the crisis. Most troubling was our finding that one third of residents had developed unfavorable opinions about recommending medicine as a career.


Our study had several limitations. First, the survey instrument has not been validated. Although instruments have been published,13 none was adequate for our purposes. Also, there is considerable variability about the definitions of quality and reports of satisfaction in residency, despite the importance of satisfaction.14,15

Second, our report addresses the effects on AGH's IM residents only. One could argue that including the perceptions of other residents and fellows at our hospital would also have been important for comparison purposes, and we agree. However, given the complexity of the reorganization within a given department, the variable responses of each program to the financial situation, the myriad alterations in faculty that each program experienced, and the enormous number of other variables that affect a postgraduate experience, we chose to focus our inquiry on a situation we could better describe. The IM residency, begun in the 1920s, is the largest of all the AGH programs, with 65 residency and 32 fellowship positions. Therefore, it could be postulated that the effects of the financial crisis on this program would find corollaries in other AGH programs.

Third, our study did not have a control group at another hospital for comparison. This comparison is important because there may be trends in perceptions, attitudes, and opinions that residency alone may affect. For example, PGY-1 residents may be more negative than PGY-3 and -4 residents due to workload or mistreatment.6 Alternatively, effects may be due to a national trend, such as the revised Health Care Financing Administration (HCFA) Evaluation and Management Guidelines,16 or an unmeasured entity such as organizational environment, which has been shown to significantly influence perceptions of teaching quality among residents.17 Maybe the residents in our study were no more disaffected than other residents regarding satisfaction, and would have been less likely to go to medical school again,18 or perhaps they would not have recommended medical school to others. However, we asked the residents to base their responses on the financial crisis at AGH, which likely had some unique qualities such as disruption of attending physician—resident relationships. We believe that there was no control group of residents at other institutions outside the financial maelstrom we experienced. And so, the data our study provides are nevertheless unique and still help define the effects of financial issues on training.

Fourth, our sample size, especially within each PGY group, was small. As such, the relative effects of some measurements may be overemphasized. In addition, conclusions drawn from statistical differences in observational studies such as this must be considered carefully. Finally, we did not use convenience samples. Despite these limitations, our study is important, because as of the writing of this report, we knew of no other study of the effects of financial instability on residents' training experiences. Further investigation of this issue is warranted, such as research into the natural history or evolution of residents' perceptions.

Our data raise several important questions for future investigation and program intervention. For example, would regular communication with and counseling sessions for residents caught in such turmoils improve training and performance? Would residents welcome training in health care systems and hospital financial issues? Would intervention improve adaptation to financial tensions that exist in many academic programs in this era of rapid health system change?


As financial crises become more frequent in the U.S. health care system, graduate medical education experience may be affected in subtle but measurable ways. The financial crisis at AGH caused an array of perceptual and educational concerns in residents. Despite the pervasive nature of these concerns, none of the residents in our study transferred. However, the long-term effects are yet to be known. Our study is important because residents are a vital component of the health care system, representing the future of academic, clinical, and research health services. Teaching hospitals must learn to assess and address residents' concerns if their institutions are caught in financial crises, by addressing morale, fellowship concerns, and changes in faculty. Residents serve as program ambassadors, and their experiences may influence recruitment and retention, a fact that has implications for residency programs, health care systems, and potentially, the future physician workforce.


1. Inglehart JK. Health policy report, rapid changes for academic medical centers. N Engl J Med. 1994;331:1391–5.
2. Anderson GF, Greenberg G, Lisk CK. Academic health centers: exploring a financial paradox. Health Aff. 1999;18:156–67.
3. Reinhart UE. The rise and fall of the physician practice management industry. Health Aff. 2000;19:42–55.
4. Tierney WM, Dexter PR, Eckert GJ, Zhou X. Effect of physician turnover on patient satisfaction in an academic primary care practice. J Clin Outcomes Manage. 2000;7:33–8.
5. Daugherty SR, Baldwin DC, Rowley BD. Learning satisfaction and mistreatment during medical internship. JAMA.1998;279:1194–9.
6. Valente E, Wyatt SM, Moy E, Levin RJ, Griner PF. Market influences on internal medicine residents' decisions to subspecialize. Ann Intern Med. 1998;128:915–21.
7. Miller RS, Dunn MR, Richter TH, Whitcomb ME. Employment—seeking experiences of resident physicians completing training during 1996. JAMA. 1998;280:777–83.
8. Wray JL, Sadowski SM. Defining teaching hospitals' GME strategy in response to new financial and market challenges. Acad Med. 1998;73:370–9.
9. Bowen JL. Adapting residency training, training adaptable residents. West J Med. 1998;168:371–7.
10. Burns LR, Cacciamani J, Clement J, Aquino W. The fall of the house of AHERF: the Allegheny bankruptcy. Health Aff. 2000;19:7–41.
11. Lyttle CS, Levey GS. The national study of internal medicine manpower: XX. the changing demographics of internal medicine residency training programs. Ann Intern Med. 1994;121:435–41.
12. Results of the National Study for Graduate Education in Internal Medicine, 〈〉. Accessed 10/4/02.
13. Bellini L, Shea JA, Asch DA. A new instrument for residency program evaluation. J Gen Intern Med. 1997;12:707–10.
14. Klessig JM, Wolfsthal SD, Levine MA, et al. A pilot study to define quality in residency education. Acad Med. 2000;75:71–3.
15. Kapur N, Appleton K, Neal RD. Sources of job satisfaction and psychological distress in GPs and medical house officers. Fam Pract. 1999;16:600–1.
16. Fihn SD, Schleyer AM, Kelly—Hedrick H, Martin DB. Effects of revised HCFA guidelines on inpatient teaching. J Gen Intern Med. 2000;15:451–6.
17. Probst JC, Baxley EG, Schell BJ, Cleghorn GD, Bogdewic SP. Organizational environment and perceptions of teaching quality in seven South Carolina family medicine residency programs. Acad Med. 1998;73:887–93.
18. Burdi MD, Baker LC. Physicians' perceptions of autonomy and satisfaction in California. Health Aff. 1999;18:134–45.
© 2003 Association of American Medical Colleges