The influence of payment method on patterns of physician practice: experience at a Canadian academic health center.

In Canada, as in most developed nations, the last decade has witnessed strenuous efforts to curtail rising health care costs. One approach to cost containment focuses on changing the way physicians are paid. In 1994 Queen's University, its hospital partners in Kingston, Canada, and clinical faculty joined this trend by initiating the Alternative Funding Plan (AFP) and forming the South Eastern Ontario Academic Medical Organization (SEAMO) by which to administer it. Previously, clinical faculty billed the provincial health plan, which provides universal publicly-funded, first-dollar coverage for the provision of all medically necessary care. Under the new plan, the health science center received from the Ontario Ministry of Health a yearly sum fixed over five years, from which clinical faculty were paid for all care. Allocation to individual departments generally reflected the relative size of previous fee-for service (FFS) billings, while each clinician negotiated yearly remuneration with a department head based largely on previous billings but adjusted in some cases to reflect of contribution to the group's practice, research and teaching responsibilities.

Since very few services are provided outside the Provincial medical plan, under the AFP the volume of a faculty member's clinical activity would be effectively de-linked from income. Moreover, that income would remain largely static over the life of the contract. This alteration in financial incentives, according to planning documents (Queen's Health Policy Research Unit, 1996), was intended to create a better match between patient need and physician encounter. By eliminating marginally-necessary care, it was anticipated that faculty would have additional time for research, educational activities, or enhanced patient care.

Did the new payment system achieve its intended clinical objective? There was a decline in the volume of clinical activity within SEAMO which was generally similar to trends throughout the province between 1992 and 1996 (Shaw, Jackson & Farquhar, 1999; Stanton, 1999). Whether this volume reduction represented the elimination of least necessary care was investigated in a pilot study using gastrointestinal endoscopy as a sentinel procedure. The volume of endoscopic procedures declined by 34 percent when the two year periods before and after initiation of the plan were compared. Significantly, the proportion of least significant post-procedure diagnoses, so-classified by faculty endoscopists, decreased. This study concluded that the results had been selectively achieved: the decrease in volume of endoscopy preferentially eliminated the least necessary care (Stanton & Parent, 1999). The present study builds on this work to explore whether the AFP, as its founders hoped, created change in the volume and mix of surgical cases in selected areas and compares any change to patterns of practice at the other academic health centers.

BACKGROUND

FFS payment encourages an increase in the volume of medical services compared to other payment formats (Hickson, Altemeier & Perrin, 1987; Broomberg & Price, 1990; Hemenway, Killen, Cashman, Parks & Bickness, 1990; Hohlen et al., 1990; Nguyen & Derrick, 1997; Retchin, 1997; Yip, 1998). Financial incentives may also alter the approach to clinical practice. Surgeons paid by FFS, in comparison to salaried practitioners (Wilson & Longmike, 1978) or participants in prepaid plans (Luke & Thomson, 1980), have been found to have different approaches to resource use and timing in the management of common surgical conditions. Similarly, bundling all components of payment for bypass surgery into a negotiated global sum has been shown to significantly change surgeon practice with respect to the use of nursing staff, laboratory tests, and intensive care beds (Cromwell, Dayhoff & Thoumaian, 1997).

Not only has remuneration been demonstrated to influence practice volume and methods, but it also has an impact on patterns of care provision. Some of these remuneration-related changes may represent the selective elimination of least necessary care. When physicians under prepaid plans were compared to FFS surgeons, the prepaid groups performed fewer procedures for which there was a large discretionary component, such as hysterectomies, tonsillectomies-adenoidectomies (LoGerfo, Efird, Diehr & Richardson, 1979), or caesarian sections (Price & Broomberg, 1990). A similar observation has been made when the cesarean rate under Medicare reimbursement is compared to the higher rate under more lucrative reimbursement from private insurance in the United States (Gruber, Kim, & Mayzlin, 1999; Keeler & Brodie 1993). A study of gynecologists who switched from FFS to a capitated contract found service volume decreased largely due to a reduction in elective procedures (Ransom, Do & McNeeley, 1996). Similarly, when physicians paid by capitation who had an equity interest in utilization incentives paid to their group were compared to those paid by traditional indemnity insurance, it was found that capitated patients in the 25 to 44 age group were admitted to hospital at a significantly lower rate for low-acuity medical conditions (Josephson & Karez, 1997).

In summary, there is a consensus that FFS practice, when compared to salary or capitation methods of payment, encourages high volume practice. Shifting an academic health center (AHC) to block funding, therefore, might be expected to result in lower volumes of service provision. What is less clear from the literature, however, is how this diminished volume of services will be distributed; that is, will all types of services decline, or will there be a preferential diminution in least necessary care?

OBJECTIVES

The objective of this study was to determine if there was a change in clinical practice patterns associated with the implementation of the AFP. The specific questions that this study addressed were:

* Was there a change in the volume of selected procedures in SEAMO associated with the implementation of the AFP?

* Was there a change in the procedural case-mix of the specialists' clinical practice in SEAMO associated with the implementation of the AFP?

* Do any changes identified over the study period in SEAMO differ from changes noted at the four other Ontario AHCs?

* What factors are identified by clinicians at SEAMO and the four other centers as being most relevant in explaining any changes noted?

METHODS

Study Design

This retrospective study took advantage of a naturally occurring experimental setting. It used a pre-post approach to analyze administrative data for 2.5 years before and 2.5 years after the AFP was implemented. In addition, survey data was collected from clinicians to assist in the interpretation of the statistical findings.

Sentinel Procedures

The study procedures (and ICD-9 codes) were: prostatectomy (72.1-72.4, 72.52, 72.59), non-obstetrical dilatation and curettage (D & C) (81.09), total hip replacement (93.51, 93.59), total knee replacement (93.41, 93.47) and permanent pacemaker insertion (49.72, 49.73). All but the latter were identified in the medical literature as having a relatively high degree of discretion in the decision to use them and, therefore, would be potentially sensitive to organizational changes (Gibson, 1997). In contrast, pacemaker insertion was characterized as much less discretionary and was used as a control.

Setting

Data were collected on the practice of clinicians at the five medical schools in Ontario (London, Hamilton, Toronto, Kingston, and Ottawa) and the 17 acute care general teaching hospitals, offering secondary through quaternary care, which are affiliated with them. SEAMO is distinguished from the other four AHCs in Ontario only by the fact that it is the smallest and that the majority of its physicians opted for a funding plan that offered a different form of remuneration. All centers operate in an environment defined by the policies of the Ministry of Health and Long-Term Care for most matters, and the Ministry of Training, Colleges and Universities for some educational activities.

Study Population

The study included every Inpatient and Same Day Surgery patient discharge record that:

1. occurred between January 1992 and December 1996 (inclusive)

2. was performed in an acute care institution identified as being affiliated with an AHC; and

3. was provided by one of the identified specialties.

Data Collection/Manipulation

Data, provided by the Provider Services Branch of the Ontario Ministry of Health, originated from the Canadian Institute of Health Information's (CIHI's) Discharge Abstract Database. Variables extracted, calculated or constructed included quarterly counts of institutions, encounters and procedures (all, selected, and diagnostic procedures), ratios of admission types (i.e., emergency vs. elective), admission source (Inpatient [InP] vs. Same Day Surgery [SDS]), and gender (female vs. male). Averages were calculated for the number of clinicians per institution, length of stay (LOS), patient age, patient severity, and number of complications. The data was acquired in two stages. The first data set was used to identify the specialties under which the selected procedures were performed since different specialties may provide a specific procedure in different centers. This subset of data encompassed all of the procedures performed during the second and fourth quarters of each year in the five year study period (1992-1996).

A second data set was obtained following confirmation of the clinical specialties responsible for performing the study procedures. It included all of the procedures performed (both InP and SDS) by the identified specialties during the full five-year period. The total number of procedures performed during each encounter was calculated, as was the total number of diagnoses per patient encounter. Data retained for responding to the study questions was then limited to include only the first three procedures performed during each patient encounter. Previous investigators have suggested that the reliability of procedures recorded in the administrative database may decline after the third listed procedure per admission (Williams & Young, 1996). These procedures accounted for over 90 percent of the workload in 3 specialties, and over 78 percent in cardiology.

In order to explore perceptions of the impact on practice patterns of factors other than the type of payment mechanism at AHCs, clinicians from the specialties from which the sentinel procedures were drawn were asked to participate in a survey. The questionnaire was developed with the assistance of one SEAMO clinician from each of the specialties. Key informants for each specialty were identified in each AHC either to provide names and mailing addresses for the relevant clinicians at their center, or to distribute the questionnaire if the former information was considered confidential. A single repeat mailing was done. All responses were anonymous and no incentives were offered for participation, other than receipt of the survey results.

Analysis

Preliminary descriptive analyses were performed to define clinical activities in each center. This and subsequent analysis used aggregate quarterly measures which were calculated as proportions, averages, ratios, or totals (as indicated for analysis) to deal with concerns that differences would result simply from the large numbers or from the discrepancies in size among the centers. Repeated measures analysis was then used to deal with the lack of independence that existed between the quarterly measures within a period. The first step in the analysis involved a bivariate analysis to define statistically significant changes that occurred in SEAMO between the pre- and post periods. This analysis was followed by multivariate, backwards stepwise repeated measures analysis to determine the statistical significance of differences found in SEAMO to those found in the other AHCs. Finally, multivariate repeated measures analysis was performed to assess the statistical significance of other factors that may have influenced any difference in change found between locations. The analysis forced the "fixed" factors (study period, AHC group, and the interaction between the two) into the model and then used a backwards, stepwise approach to determine the significance of the other factors on the proportion of workload devoted to the sentinel procedures.

Chi-square analysis was performed to assess the significance of differences between SEAMO and the other academic centers. Differences were assessed overall and by specialty when sufficient numbers were available.

RESULTS

Describing the Clinical Environment

Exhibit 1 displays the volume of clinical activity in SEAMO and the other AHCs. It should be noted there was a large discrepancy in the volume of activities between the groups. While the average number of specialists per institution was comparable for orthopedics and urology, there was a much higher average number of cardiologists and obstetrician/gynecologists per institution in the other AHCs than in SEAMO. To control for the discrepancy in size, proportional, percentage, and/or ratio measures were used whenever possible in the analysis.

Changes in Practice Profile in SEAMO by Specialty

1. Cardiology/Cardiac Surgery:

A significant increase in the volume, from 2,582 to 2,930, of SDS admissions for Cardiologists / Cardiac Surgeons was accompanied by a significant decrease in the average number of procedures performed during an encounter and in the severity of patient illnesses as measured by the average number of diagnoses (Exhibit 2). These changes in the SDS activities of Cardiologists / Cardiac Surgeons coincided with a slight, but statistically significant, increase in the proportion of permanent pacemakers implanted on a SDS basis. One significant change found in Inpatient activity--a decrease in the ratio of emergency to elective admissions--was reflected in a similar decrease in overall activity. This change was accompanied by a statistically significant increase in Inpatient severity (Exhibit 2).

2. Obstetric/Gynecology:

Exhibit 2 also shows there was a significant decrease in the volume of both Inpatient and SDS admissions for Obstetricians / Gynecologists. While there were 995 fewer Inpatients admitted during the post study period, these patients were `sicker' in that they had an average of 2.5 more diagnoses recorded and an average 1.74 more procedures performed during each encounter. These changes occurred despite a significant decrease in the emergency to elective admission ratio. During the same time SDS experienced a drop of 407 admissions which was accompanied by a slight, but statistically significant, reduction in the average number of patient diagnoses and procedures performed during each encounter and by an increase in the ratio of emergency to elective admission ratio.

3. Orthopedics

There was a statistically significant increase in the ratio of Orthopedic surgeons' Inpatient to SDS admissions (Exhibit 2). Along with this increase was a decrease in the average length of patient stay despite a significant increase in patient severity. The increase in Inpatient severity was reflected in a similar increase overall, despite a significant decrease in the severity of patients admitted on a SDS basis.

4. Urology

Urologists saw a significant decline in their Inpatient to SDS admission ratio during the study period (Exhibit 2). The decrease in Inpatient admissions was accompanied by decreases in both the ratio of emergency to elective admissions and the proportion of procedural workload attributed to performing prostatectomies. Coinciding with the decline in SDS admissions was an increase in the ratio of female to male patients.

Overall Changes at SEAMO vs. other AHCs

Analysis found no significant difference between the pre and post changes in the volume of the specialties' procedural workload in SEAMO and those in the other AHCs (Exhibit 3). Despite the lack of difference in volume changes, however, some significant differences in clinical activities were detected in the Inpatient clinical practice of two specialties, Obstetrics/Gynecology and Urology. In the other AHCs, Obstetrician/Gynecologists reduced the proportion of their workload devoted to non-Obstetrical D&Cs (from 1.0 to 0.6 percent), while in SEAMO, clinicians under the same specialty increased the proportion of their inpatient workload devoted to nonobstetrical D&Cs from 2 to 3 percent. SEAMO Urologists reduced the proportion of their workload devoted to inpatient prostatectomies by 3 percent (from 18 to 15 percent), but the clinical practice of their counterparts in the other AHCs remained stable with 14 percent devoted to prostatectomies.

Adding other Factors to the Analysis

To assess some of the other factors that may have influenced, or changed with, the differing change of proportion of specialists' workload in SEAMO, a backwards, stepwise repeated measures analysis with replacement was used. The final models included all of the variables that significantly influenced the proportion of the selected procedures performed (dependent variable) for each specialty, as well as the `fixed' factors.

1. Factors Relating to Differences in the Obstetrician/Gynecologists Practices

Along with the fixed factors, the final model for the proportion of the Inpatient clinical workload of Obstetrician/Gynecologists that was devoted to non-obstetrical D&Cs included the interactions of the study period and AHC Group with each of the following: volume of services provided, the ratio of emergency to elective admissions, the average number of procedures performed during each encounter, the average number of diagnoses recorded for each patient, and the proportion of patients serviced who were from `outside' the AHC's care area (Exhibit 4).

2. Factors Relating to Differences In the Urologists Practices

The model of the proportion of Urologists' Inpatient clinical workload devoted to prostatectomies also included the `fixed' factors. Significant factors in the model were the interaction between the study period and AHC Group and each of the following variables: the volume of services provided, the average number of diagnoses per patient, the average length of patient stay, and the proportion of elderly (65 years of age and over) patients (Exhibit 4).

Clinician Perceptions of Practice Change

Analysis of the questionnaire sent to all clinicians (n = 369) working under the defined specialties in an Ontario AHC focused on whether the clinicians had changed their clinical practice and, if so, for what reasons. Of the 23 SEAMO clinicians who responded to the question about whether they had changed their clinical practice, 30 percent (n = 7) reported that they had changed their approach to the selected procedure. Of those who changed their clinical practice, the vast majority (n = 6) stated they had made the change because of scientific evidence. The two clinicians who had reported changes related to factors other than scientific evidence were both Urologists and stated resource changes were the basis for their practice change.

Similarly, 32 of the 144 (22 percent) clinicians in the other AHCs who responded to this question reported having changed their clinical practice of the selected procedure. While 29 of these clinicians related the change to scientific evidence, eight of the clinicians cited resource changes and `other' factors as contributing to the change. Exhibit 5 displays overall responses to the questions; a breakdown by specialty could not be provided due to confidentiality issues. Chi-square analysis revealed no significant differences in the responses between SEAMO and the other AHCs.

DISCUSSION

This study found no significant overall change in the volume of selected procedures and a minimal change in the procedural case-mix, confined to inpatients, of study specialists at SEAMO following implementation of the AFP. These practice patterns were for the most part similar to those seen at other AHCs in Ontario where clinicians were paid on a FFS basis. Thus, the study has failed to support the hypothesis that switching to global funding at an AHC will create financial incentives which favor decreased volume of service, or that the decrease will be achieved by selectively reducing discretionary interventions. There are a number of possible explanations for these findings.

The study design may have inhibited the identification of change. It is possible that established physician practice habits require more than the 2.5 years post-AFP examined in this study to show a significant change, though the American literature does suggest that responses often occur rapidly when economic incentives are changed. It is possible that the choice of other sentinel procedures would have produced different results. However, since the conditions chosen for study were selected on the basis of a literature review that identified them as both frequently performed and relatively discretionary, it seems unlikely that less discretionary procedures would have shown greater change. Finally, as the survey of clinicians revealed, many believed they changed their practice patterns in response to new clinical guidelines or evidence that became available during the study period. Adoption of evidence-based practices would tend to be evenly distributed across all AHCs in the province, thereby diminishing differences in practice patterns between centers.

Two characteristics of the study site may have diminished the detection of differences between it and other academic health centers. First, it is possible that the AFP represented very little change from the previous payment system at SEAMO. When full-time University faculty billed the provincial plan on a FFS basis, their incomes were subject to ceilings imposed by the University, thereby attenuating the normal volume incentives associated with FFS. However, this hypothesis ignores three important points. Physicians were allowed to use FFS earnings beyond their ceiling to cover all tax-deductible expenses of practice. As well, a consistently large income above the set ceiling was a critical bargaining point in securing a higher ceiling. Finally, physicians gained stature within their departments by contributing large amounts of over-ceiling income. These points suggest the ceiling system shared income-maximizing incentives with the regular FFS system which differed significantly from the AFP.

A second site characteristic which may have diminished the impact of a change in remuneration is that clinicians affiliated with SEAMO have a virtual monopoly on the provision of specialty care for the region. Even for discretionary procedures, there are almost no other specialists to whom care can be relegated, thereby inhibiting SEAMO clinicians from readily altering their practice patterns. However, for patients from the periphery of the catchment area, specialty care in contiguous AHCs would be a feasible alternative. This market dominance and the fact that SEAMO is the smallest of the Ontario AHCs, are the only features known to distinguish the study center form the comparison group.

It is possible that physician characteristics diminished the response to changes in remuneration method. Since individual incomes under the AFP initially changed little from those achieved under the previous FFS arrangement, physicians may have simply felt most comfortable with maintaining existing practice patterns. Balanced against this interpretation are the results of a faculty survey from 1996 which indicated that 61 percent of respondents were eager to spend more time on research, with a corresponding reduction in clinical load (Cosby & Middleton, 1996). However, since responses were not coded by department, it is possible that the surgeons responsible for the procedures reviewed in this study did not share this consensus.

Finally, the broader health system environment in Ontario may have influenced study results. The literature predicting a volume-reduction response to a non-FFS environment is almost exclusively based on American experience; however, the system of publicly-funded, universal health care in Canada may create an environment in which physicians respond quite differently to economic incentives. For example, the two relevant Canadian studies dealing with the response to alterations in medical fees (Labelle Hurley & Rice, 1990; Hurley & Labelle 1995) failed to report physician behavioral changes similar to those in the United States. Similarly, unlike their American counterparts, Ontario physicians participating in HMO-like practice settings did not reduce hospitalization rates of their patients despite financial incentives to do so (Hutchison Birch, Hurley, Lomas & Stratford-Devai, 1996). To date a single Canadian study has linked fee pooling by a group of specialists with a diminution in clinical service (Bland, Oppenheimer, Holmes & Wen, 2001).

During the study period, physicians in Ontario were subject to a professional fee cap, beyond which the province discounted the fees paid. The capping policy began in April 1991 and continued with minor changes until April 1996 when specialty-specific caps were introduced (Dowdall & Ramchandar, 1999). It is possible that FFS physicians at other AHCs reacted to fee caps by reducing their volume of practice through the elimination of discretionary cases. This would tend to obscure any difference between their practices and physicians at SEAMO. It is questionable, however, how significant an impact the fee caps had on the financial incentives of clinicians at Ontario AHCs since the discounted fees would be applicable largely to the portion of the clinicians' billings which would have been returned to the their medical school. In addition, clinicians providing highly specialized services in their field were eligible to apply for an exemption from the billing thresholds and billing for some services, including pacemaker insertion, was excluded from threshold calculations (Ontario Medical Association [not dated]).

Lastly, a number of changes occurred in the Ontario health system from the late 1980s through the period covered by this study. For example, the number of hospital separations, patients days and length of stay, all fell steadily from a high in 1987-1988 to 1994-1995, the year in which the AFP was initiated (Anderson, Chart, Carter & Axcell, 1996). While it is likely that teaching hospitals were comparatively protected in this process, it is possible that resources were sufficiently constrained as to have already significantly reduced episodes of discretionary care across the province. If this were the case, it would serve to diminish detectable differences in practice between the FFS centers and the AFP.

CONCLUSION

While further research, covering a longer period of time and capturing a wider range of surgical and medical services, is required, at present it is reasonable to conclude that the AFP had few discernable implications for clinical practice patterns. The current study suggests under global budget payment there will be no appreciable decrease in discretionary surgical procedures known to be costly and commonly done. Hospital managers may find the cost of supplying surgical services is independent of remuneration systems for physicians and that, because practice patterns remain static, there is no need to plan for substitute services. Further, the negative results in this study suggest surgical personnel needs, including specialty mix, required to service a given population are independent of remuneration format. This finding is of importance to medical educators and health service planners.

EXHIBIT 1
DESCRIBING THE CLINICAL ENVIRONMENT:
VOLUME OF DISCHARGES AND NUMBER OF SPECIALISTS PER INSTITUTION
1992-1997

                                                  SEAMO    Other AHCs
Cardiology
  Inpatient                                        5,590      117,055
  Same Day Surgery (SDS) *                         5,512       26,842
  Average No. Specialist/Institution                   6           11

Obstetrics/Gynecology
  Inpatient                                       12,641      218,427
  Same Day Surgery (SDS) *                         9,099      164,459
  Average No. Specialist/Institution                   9           15

Orthopedics
  Inpatient                                       16,290      168,256
  Same Day Surgery (SDS) *                        10,093       63,930
  Average No. Specialist/Institution                   7            6

Urology
  Inpatient                                        4,400       58,079
  Same Day Surgery (SDS) *                        11,922      140,427
  Average' No. Specialist/Institution                  4            4

EXHIBIT 2
CLINICAL PRACTICE CHARACTERISTICS WITHIN SEAMO

                                                      OVERALL

PANEL A                                        Pre-             Post-
                                                AFP               AFP

Cardiologists/Cardiac Surgeons
  Volume of Admissions                      5,393          5,709
  Inpatient to SDS Ratio                        1.09           0.95
  Average LOS                                   4.94           4.68
  Ave. No. of Procedures/Encounter              3.29           3.23
  Ave. No. of Diagnoses/Patient                 3.15           3.20
  Emergency vs. Elective Admissions             0.31           0.24 ***
  Proportion of procedures:                     0.02           0.02
  Pacemakers
  Proportion of elderly patients                0.43           0.44
  Proportion of patients with
  complications                                 0.24           0.21
  Ratio of Female to Male patients              0.45           0.49
Obstetricians/Gynecologists
  Volume of Admissions                     11,641         10,175 ***
  Inpatient to SDS Ratio                        1.42           1.35
  Average LOS                                   2.86           2.72
  Ave. No. of Procedures/Encounter              3.33           4.18 **
  Ave. No. of Diagnoses/Patient                 3.43           4.72 **
  Emergency vs. Elective Admissions             0.07           0.07
  Proportion of procedures:                     0.05           0.04
  Non-OBSs D&Cs
  Proportion of elderly patients                0.05           0.05
  Proportion of patients with
  complications                                 0.17           0.16
  Ratio of Female to Male patients              --               --

PANEL B

Orthopedic Surgeons
  Volume of Admissions                     13,230         13,153
  Inpatient to SDS Ratio                        1.53           1.73 ***
  Average LOS                                   6.15           5.74
  Ave. No. of Procedures/Encounter              2.24           2.27
  Ave. No. of Diagnoses/Patient                 2.90           3.11 **
  Emergency vs. Elective
  Admissions                                    0.33           0.31
  Proportion of procedures:
  Hip Replacements                              0.04           0.04
  Proportion of procedures:
  Knee Replacements                             0.04           0.04
  Proportion of elderly patients                0.26           0.29
  Proportion of patients with
  complications                                 0.34           0.35
  Ratio of Female to Male patients              0.97           0.95
Urologists
  Volume of Admissions                      8,862          7,460 ***
  Inpatient to SDS Ratio                        0.39           0.35 ***
  Average LOS                                   3.49           3.17
  Ave. No. of Procedures/Encounter              1.52           1.51
  Ave. No. of Diagnoses/Patient                 3.25           3.16
  Emergency vs. Elective
  Admissions                                    0.08           0.06 ***
  Proportion of procedures:
  Prostatectomies                               0.07           0.05 ***
  Proportion of elderly patients                0.43           0.40
  Proportion of patients with
  complications                                 0.07           0.08
  Ratio of Female to Male patients              0.29           0.33 **

                                                 INPATIENT
PANEL A
                                            Pre-           Post-
                                             AFP            AFP

Cardiologists/Cardiac Surgeons
  Volume of Admissions                      2,811          2,779
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   8.64           8.59
  Ave. No. of Procedures/Encounter              3.65           3.70
  Ave. No. of Diagnoses/Patient                 4.90           5.37 ***
  Emergency vs. Elective Admissions             0.86           0.67 **
  Proportion of procedures:                     0.03           0.03
  Pacemakers
  Proportion of elderly patients                0.47           0.47
  Proportion of patients with
  complications                                 0.45           0.41
  Ratio of Female to Male patients              0.44           0.48
Obstetricians/Gynecologists
  Volume of Admissions                      6,818          5,823 ***
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   4.17           4.00
  Ave. No. of Procedures/Encounter              4.41           6.15 ***
  Ave. No. of Diagnoses/Patient                 4.97           7.42 ***
  Emergency vs. Elective Admissions             0.11           0.08 **
  Proportion of procedures:                     0.02           0.03
  Non-OBSs D&Cs
  Proportion of elderly patients                0.07           0.07
  Proportion of patients with
  complications                                 0.28           0.26
  Ratio of Female to Male patients             --             --

PANEL B

Orthopedic Surgeons
  Volume of Admissions                      7,953          8,337
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   9.56           8.50 **
  Ave. No. of Procedures/Encounter              2.60           2.58
  Ave. No. of Diagnoses/Patient                 3.78        4.07 **
  Emergency vs. Elective
  Admissions                                    0.67           0.59
  Proportion of procedures:
  Hip Replacements                              0.05           0.06
  Proportion of procedures:
  Knee Replacements                             0.06           0.06
  Proportion of elderly patients                0.35           0.37
  Proportion of patients with
  complications                                 0.59           0.55
  Ratio of Female to Male patients              1.07           1.01
Urologists
  Volume of Admissions                      2,478          1,922 ***
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   7.39           7.02
  Ave. No. of Procedures/Encounter              2.29           2.35
  Ave. No. of Diagnoses/Patient                 7.35           7.82
  Emergency vs. Elective
  Admissions                                    0.36           0.29 ***
  Proportion of procedures:
  Prostatectomies                               0.18           0.15 ***
  Proportion of elderly patients                0.55           0.51
  Proportion of patients with
  complications                                 0.18           0.22
  Ratio of Female to Male patients              0.22           0.26

                                                        SDS

PANEL A                                         Pre-          Post-
                                                AFP            AFP

Cardiologists/Cardiac Surgeons
  Volume of Admissions                      2,582          2,930 ***
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   1              1
  Ave. No. of Procedures/Encounter              2.91           2.78 **
  Ave. No. of Diagnoses/Patient                 1.28           1.14 ***
  Emergency vs. Elective Admissions             0.00           0.00
  Proportion of procedures:                     0.00           0.01 ***
  Pacemakers
  Proportion of elderly patients                0.38           0.43
  Proportion of patients with
  complications                                 0.01           0.01
  Ratio of Female to Male patients              0.47           0.50
Obstetricians/Gynecologists
  Volume of Admissions                      4,753          4,346 **
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   1.00           1.00
  Ave. No. of Procedures/Encounter              1.77           1.56 ***
  Ave. No. of Diagnoses/Patient                 1.25           1.16 ***
  Emergency vs. Elective Admissions             0.00           0.03 ***
  Proportion of procedures:                     0.08           0.05 ***
  Non-OBSs D&Cs
  Proportion of elderly patients                0.02           0.02
  Proportion of patients with
  complications                                 0.01           0.01
  Ratio of Female to Male patients             --             --

PANEL B

Orthopedic Surgeons
  Volume of Admissions                      5,277          4,816
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   1.00           1.00
  Ave. No. of Procedures/Encounter              1.69           1.74
  Ave. No. of Diagnoses/Patient                 1.56           1.45 ***
  Emergency vs. Elective
  Admissions                                    0.01           0.01
  Proportion of procedures:
  Hip Replacements                              0.00           0.00
  Proportion of procedures:
  Knee Replacements                             0.00           0.00
  Proportion of elderly patients                0.13           0.15
  Proportion of patients with
  complications                                 0.00           0.00
  Ratio of Female to Male patients              0.85           0.86
Urologists
  Volume of Admissions                      6,384          5,538 ***
  Inpatient to SDS Ratio                       --             --
  Average LOS                                   1.00           1.00
  Ave. No. of Procedures/Encounter              1.22           1.22
  Ave. No. of Diagnoses/Patient                 1.66           1.55
  Emergency vs. Elective
  Admissions                                    0.00           0.00
  Proportion of procedures:
  Prostatectomies                               0.00           0.00
  Proportion of elderly patients                0.35           0.34
  Proportion of patients with
  complications                                 0.00           0.00
  Ratio of Female to Male patients              0.35           0.38 **

*** p [less than or equal to] 0.01

** p [less than or equal to] 0.05

EXHIBIT 3
ASSESSING DIFFERENCE IN THE CHANGE OF CHEMICAL ACTIVITY BETWEEN
SEAMO AND OTHER AHCs

                                             F STATISTIC
                                              (P-VALUE)

                                   Overall    Inpatient      SDS

Cardiology
                                    0.33        0.14         0.94
  Volume                           (0.57)      (0.71)       (0.33)

  Proportion of Workload:           0.00        1.51         0.33
  Permanent Pacemakers             (0.96)      (0.22)       (0.57)

Obstetrics/Gynecology
                                    0.04        0.22         0.00
  Volume                           (0.84)      (0.64)       (0.99)

  Proportion of Workload:           0.08        9.48         0.37
  non-Obstetrical D&C              (0.77)      (0.00) ***   (0.55)

Orthopedics
  Volume                            0.19        0.01         3.16
                                   (0.67)      (0.90)       (0.08)

  Proportion of Workload: Total     1.96        0.64
  Hip Replacement                  (0.16)      (0.42)         --

  Proportion of Workload:           0.19        0.83
  Total Knee Replacement           (0.67)      (0.36)         --

Urology
  Volume                            0.01        0.67         0.07
                                   (0.92)      (0.41)       (0.79)

  Proportion of Workload:           0.18       10.10         0.49
  Prostatectomy                    (0.67)      (0.00)***    (0.49)

*** p [less than or equal to] 0.01

** p [less than or equal to] 0.05

EXHIBIT 4
PROPORTION OF WORKLOAD FOR SENTINEL PROCEDURES AT SEAMO IN
TWO SPECIALTIES

OBS/GYN
Proportion of workload devoted to                    F            P
non-obstetrical D&Cs depending upon:             Statistic      Value

  Period                                            0.14       0.72
  AHC Group                                         4.67       0.00 ***
  AHC Group * Period                                0.69       0.41
  Volume * AHC Group * Period                      11.1        0.00 ***
  Emergency vs. Elective Admissions
    * AHC Group * Period                           17.76       0.00 ***
  Ave. No. of Procedures per Encounter
    * AHC Group * Period                            6.12       0.00 ***
  Ave. No. of Diagnoses per Patient
    * AHC Group * Period                           10.03       0.00 ***
  Proportion of patients from outside
    the AHC care area * AHC Group *                 6.88       0.00 ***
  Period

UROLOGY
Proportion of workload devoted to
prostatectomies depending upon:

  Period                                            0.28       0.60
  AHC Group                                         3.92       0.05 **
  AHC Group * Period                                0.98       0.33
  Volume * AHC Group * Period                      21.31       0.00 ***
  Ave. No. of Diagnoses per
    Patient * AHC Group * Period                    4.32       0.00 ***
  Ave. LOS * AHC Group * Period                     5.27       0.00 ***
  Proportion of elderly
    patients * AHC Group * Period                   2.90       0.03 **

*** p [less than or equal to] 0.01

** p [less than or equal to] 0.05

EXHIBIT 5
SURVEY RESPONSES FROM ALL SPECIALTIES COMBINED

PANEL A                             SEAMO

                                     75%
Surveys completed                 (n = 32)

                            % Positive Responses
                              (Total Responses)

Change in method                   78.95%
  of Remuneration?                (n = 19)
Aware of Clinical                  41.18%
  Practice Guidelines?            (n = 17)
Aware of Scientific                47.06%
  Evidence?                       (n = 17)
Quality Assurance                  83.33%
  mechanism?                      (n = 18)
Organizational Changes?            55.56%
                                  (n = 18)
Changes in Resources?              82.35%
                                  (n = 17)

PANEL A                           Other AHCs

                                      47%
Surveys completed                  (n = 337)

                             % Positive Responses
                               (Total Responses)

Change in method                     5.38%
  of Remuneration?                 (n = 93)
Aware of Clinical                   70.41%
  Practice Guidelines?             (n = 98)
Aware of Scientific                 41.11%
  Evidence?                        (n = 90)
Quality Assurance                   75.64%
  mechanism?                       (n = 78)
Organizational Changes?             53.93%
                                   (n = 89)
Changes in Resources?               60.92%
                                   (n = 87)

                                                     Number of
PANEL B                                         Responses by Category

1. Were There
  Organizational Changes?                                           +
  Direct Restriction
    To Procedures
  Staff Number                                                      9
  Volume of Personnel                                              <3
  Type of Personnel                                                 7
  Skills of Personnel                                               8
  Clinical Time Commitment of Personnel                             5
  Research Focus of Personnel                                       5
  Educational Focus of Personnel                                    8
2. Were There Resource Changes?                                     +
  Clinic Time Allotment                     [less than or equal to] 3
  OR Time Allotment
  Technology Allotment
  Technology Type                           [less than or equal to] 3
  Equipment Allotment
  Equipment Type                            [less than or equal to] 3
  Bed Allotment

                                                     Number of
PANEL B                                         Responses by Category

1. Were There
  Organizational Changes?                                           =
  Direct Restriction
    To Procedures
  Staff Number                                                      4
  Volume of Personnel                                               8
  Type of Personnel                                                 4
  Skills of Personnel                       [less than or equal to] 3
  Clinical Time Commitment of Personnel                             7
  Research Focus of Personnel                                       6
  Educational Focus of Personnel                                    4
2. Were There Resource Changes?                                     =
  Clinic Time Allotment
  OR Time Allotment
  Technology Allotment
  Technology Type
  Equipment Allotment
  Equipment Type
  Bed Allotment

                                                     Number of
PANEL B                                         Responses by Category

1. Were There
  Organizational Changes?                                           -
  Direct Restriction
    To Procedures                           [less than or equal to] 3
  Staff Number                                                      4
  Volume of Personnel                       [less than or equal to] 3
  Type of Personnel
  Skills of Personnel
  Clinical Time Commitment of Personnel
  Research Focus of Personnel
  Educational Focus of Personnel
2. Were There Resource Changes?                                     -
  Clinic Time Allotment                                             5
  OR Time Allotment                                                13
  Technology Allotment                                              4
  Technology Type
  Equipment Allotment                                               6
  Equipment Type
  Bed Allotment                                                    10

                                                    Number of
PANEL B                                         Responses by Category

1. Were There
  Organizational Changes?                                           +
  Direct Restriction
    To Procedures                            [less than or equal to] 3
  Staff Number                                                      31
  Volume of Personnel                                               25
  Type of Personnel                                                 25
  Skills of Personnel                                               24
  Clinical Time Commitment of Personnel                             20
  Research Focus of Personnel                                       20
  Educational Focus of Personnel                                    18
2. Were There Resource Changes?                                      +
  Clinic Time Allotment                                              7
  OR Time Allotment                                                  5
  Technology Allotment                                              18
  Technology Type                                                   10
  Equipment Allotment                                                9
  Equipment Type                                                     5
  Bed Allotment                              [less than or equal to] 3

                                                     Number of
PANEL B                                         Responses by Category

1. Were There
  Organizational Changes?                                           =
  Direct Restriction
    To Procedures                           [less than or equal to] 3
  Staff Number                                                     15
  Volume of Personnel                                              17
  Type of Personnel                                                18
  Skills of Personnel                                              23
  Clinical Time Commitment of Personnel                            23
  Research Focus of Personnel                                      22
  Educational Focus of Personnel                                   23
2. Were There Resource Changes?                                     =
  Clinic Time Allotment                     [less than or equal to] 3
  OR Time Allotment                         [less than or equal to] 3
  Technology Allotment                      [less than or equal to] 3
  Technology Type
  Equipment Allotment                       [less than or equal to] 3
  Equipment Type
  Bed Allotment

                                                     Number of
PANEL B                                         Responses by Category

1. Were There
  Organizational Changes?                                           -
  Direct Restriction
    To Procedures                                                  14
  Staff Number                                                     10
  Volume of Personnel                                              12
  Type of Personnel                                                 8
  Skills of Personnel                                               6
  Clinical Time Commitment of Personnel                             6
  Research Focus of Personnel                                       4
  Educational Focus of Personnel                                    6
2. Were There Resource Changes?                                     -
  Clinic Time Allotment                                            11
  OR Time Allotment                                                33
  Technology Allotment                                             15
  Technology Type                           [less than or equal to] 3
  Equipment Allotment                                              18
  Equipment Type                            [less than or equal to] 3
  Bed Allotment                                                    43

+ Increases/Improvements

= no change

- decreases

REFERENCES

Anderson, G.M., Chart, B., Carter, J.A., & Axcell, T. (1996). An overview of trends in the use of acute care hospitals, physician and diagnostic services, and prescription drugs, in V. Goel, J.I. Williams, G.M. Anderson, P. Blackstien-Hirsch. C. Fooks, C.D. Naylor (Eds.) Patterns of Health Care In Ontario, the ICES Practice Atlas. 2nd ed. Ottawa, ON: Canadian Medical Association.

Bland, E., Oppenheimer, L., Holmes P., & Wen, S. (2001). The effect of income pooling within a call group on rates of obstetric intervention, Canadian Medical Association Journal, 164,337-339.

Broomberg, J., & Price, M.R. (1990). The impact of the fee-for-service reimbursement system on the utilization of health services, part 1, a review of the determinants of doctors' practice patterns, South Africian Medical Journal, 78,130-132.

Cosby, JL, Middleton, M.J. (1996). The Perceptions of Clinical Medical Faculty towards the Alternative Funding Plan, (Technical Report), Kingston, ON, Queen's Health Policy Research Unit.

Cromwell, J., Dayhoff, D.A., & Thoumaian, A.H. (1997). Cost savings and physician responses to global bundled payments for Medicare heart bypass surgery, Health Care Financing Review, 19(1),41-57.

Dowdall, C., & Ramchandar, L. (1999). The application of physician billing thresholds in Ontario and Canada, Ontario Medical Review, 66,17-21.

Gibson, L (1997). High Volume, High Cost and Discretionary Procedures Identified In the Literature and South Eastern Ontario Health Science Centre, (Technical Report), Kingston, ON, Queen's Health Policy Research Unit.

Gruber, J., Kim, J., & Mayzlin, D. (1999). Physician fees and procedure intensity, the case of cesarean delivery, Journal of Health Economics, 18,473-490.

Hemenway, D., Killen, A., Cashman, S.B., Parks, C.L., & Bickness, W.J. (1990). Physician's responses to financial incentives--evidence from a for-profit ambulatory care centre, New England Journal of Medicine, 322,1059-1063.

Hickson, G.B., Altemeier, W.A., & Perrin, J.M. (1987). Physician reimbursement by salary or fee-for-service: effect on physician practice behavior, Pediatrics, 80,34-350.

Hohlen, M.M., Manheim, L.M., Fleming, G.V., Davidson, S.M., Yudkowsky, B.K., Werner, S.M., & Wheatley, G.M. (1990). Access to office-based physicians under capitation reimbursement and Medicaid case management--findings from the children's Medicaid program, Medical Care, 28(1),59-68.

Hurley J., & Labelle, R. (1995). Relative Fees and the Utilization of Physicians' Services in Canada, Health Economics, 4,419-438.

Hutchison B., Birch, S., Hurley, J., Lomas, J., & Stratford-Devai, F. (1996). Do physician-payment mechanisms affect hospital utilization? a study of health services organizations in Ontario, Canadian Medical Association Journal, 154,653-661.

Josephson, G.W., & Karcz, A. (1997). The impact of physician economic incentives on admission rates of patients with ambulatory sensitive conditions: an analysis comparing two managed care structures and indemnity insurance, The American Journal of Managed Care, 3(1),49-56.

Keeler, E.C., & Brodie, M. (1993). Economic incentives in the choice between vaginal delivery and cesarean section, The Milbank Quarterly, 71(3),365-404.

Labelle, R., Hurley, J., & Rice, T. (1990). Financial Incentives and Medical Practice: Evidence From Ontario on the Effect of Changes in Physician Fees on Medical Care Utilization (Working Paper), Hamilton, ON, McMaster University Centre for Health Economics and Policy Analysis.

LoGerfo, J.P., Efird, R.A., Diehr, P.K., & Richardson, W.C. (1979). Rates of surgical care in prepaid group practices and the independent setting: what are the reasons for the differences? Medical Care, 17,17.

Luke, R.D., & Thomson, M.A. (1980). Utilization of within-hospital services: a study of the effects of two forms of group practice. Medical Care, 18(2),219-227.

Nguyen, N.S., & Derrick, F.W. (1997). Physician behavioral response to a Medicare price reduction, Health Services Research, 32,283-298.

Ontario Medical Association (not dated) Threshold Levels, Exceptions (OMA Bulletin): http://www.oma.org/pcomm/faxnet/thres4.htm. Accessed June 21, 2002.

Price, M.R., & Broomberg, J. (1990). The impact of the fee-for-service reimbursement system on the utilization of health services: part iii, a comparison of caesarean section rates in white nulliparous women in the private and public sectors, South African Medical Journal, 78(4),136-138.

Queen's Health Policy Research Unit (1996). Alternative Funding Plan Evaluation Documents Model, (Technical Report), Kingston, ON: Queen's Health Policy Research Unit.

Ransom, S.B., Do, S., & McNeeley, G. (1996). The effect of capitated and fee-for-service remuneration on physician decision making in gynecology. Obstetrics and Gynecology, 87,707-710.

Retchin, S.M. (1997). The modification of physician practice patterns, Clinical Practice Management, 5(4),202-207.

Shaw, R., Jackson, A., & Farquhar, D. (1999). Waiting-times for first appointments with clinical specialists at a Canadian academic health sciences centre before and after the implementation of an alternative funding plan, Annals, Royal College of Physicians and Surgeons, 32(6).

Stanton, S. (1999). Population Health and Health Care Provision, an Overview of Trends in SEAMO, (Technical Report), Kingston, ON: Queen's Health Policy Research Unit.

Stanton, S., & Parent, K. (1999). The impact of change in physician remuneration on clinical practice---a feasibility study focusing on upper GI endoscopy, Research in Health Care Financial Management, 5(1),41-56.

Williams, J.I., & Young, W. (1996). A summary of studies on the quality of health care administrative databases in Canada. In V. Goel, J.I.Williams, G.M.Anderson, P. Blackstien-Hirsch. C. Fooks, C.D. Naylor (Eds.) Patterns of Health Care in Ontario, the ICES Practice Atlas. 2nd ed. Ottawa, ON: Canadian Medical Association.

Wilson, S.E., & Longmike, W.P. (1978). Does method of surgeon payment affect surgical care? Journal of Surgical Research, 24,457-468.

Yip, W.C. (1998). Physician response to Medicare fee reductions: changes in the volume of coronary artery bypass (CABG) surgeries in the Medicare and private sectors, Journal of Health Economics, 17,675-699.

Sally Stanton
Queen's University (Canada)

Samuel E.D. Shortt
Queen's University (Canada)

Research for this study was made possible by a grant from the Canadian Health Services Research Foundation, and by salary support from by the South Eastern Ontario Academic Medical Organization.

Address for correspondence: S.E.D. Shortt, Centre for Health Services & Policy Research, Queen's University, Kingston, ON K71 3N6 Canada, [e-mail].