|
...control goal divergence implementing information systems to better monitor agents' behavior and outcomes. But what happens when the principal does not possess legitimacy to impose an agent's use of information and/or behavioral conformance? This study investigates this situation. Through an action research project, a physicians' profiling system (PPS) was used to monitor and benchmark physicians' clinical practices and outcomes resulting in changed practice behaviors in closer congruence with management's goals.
The PPS project represents a successful attempt of a hospital's management (principal) to "informate the clan" of physicians (agents) to reduce clinical procedural costs and adopt practices benchmarked to produce better outcomes. This research moves beyond directly controlling informated workers through legitimized managerial authority to a better understanding of how to informate autonomous professionals. Emerging insights suggest that a clan can be informated if the principal can improve the perceived legitimacy of the information (the message), legitimize the technical messenger (customized user interface), legitimize the human messenger (boundary spanners and influential clan members), and facilitate an environment where clan-based discussion, using the information provided by the principal, is incorporated into the process of concertive control.
Keywords: Action research, informating, clan, control and IT implementation, IT-based performance monitoring, agency theory, concertive control, health care information systems
**********
Introduction
Control has long been a centerpiece concept in management literature (Eisenhardt 1985; Govindarajan and Fisher 1990; Ouchi 1979; Thompson 1967). Control typically is viewed through the formal and informal performance evaluation processes by which behaviors and outcomes are measured, evaluated, and rewarded. Control theories have been the basis for research in information systems (IS), including changes in behavioral rules and procedure of control (Orlikowski 1991; Sia and Neo 1997), the role of self managed teams and clan-like mechanisms (Henderson and Lee 1992; Klein and Kraft 1994), and self monitoring in controlling IS development projects (Kirsch 1996, 1997). Theories of control, including bureaucratic, economic, clan, and self-control, all see information availability and accuracy concerning behaviors and outcomes as key factors in shaping the structure of an organization's control systems. Information systems play a central role in making behaviors and outcomes more transparent between parties. In her classic work In the Age of the Smart Machine: The Future of Work and Power, Shoshanna Zuboff (1988) characterized this phenomenon as informating. She states that information technology (IT) not only automates,
but simultaneously generates information about the underlying productive and administrative processes through which an organization accomplishes its work: It provides a deeper level of transparency to activities that had been either partially or completely opaque.... Activities, events and objects are translated into and made visible by information when a technology infomates (pp. 9-10).
While the above-mentioned studies (and many others in this line of inquiry) offer a valuable cumulative tradition concerning the use of IS in situations where the controller has legitimate authority over the controlled, they do not directly address the situation where the controller, attempting to informate workers' behaviors, does not possess such legitimacy. This paper addresses such a problem situation. Specifically, this study tracks a hospital's attempts to exercise cost and outcome control over physicians via an information system by informating physicians' practice decisions with performance information. An initial direct informating attempt by management was viewed as a failure because, while it resulted in greater transparency, it failed to result in significant behavioral change in the clinical practice of the physicians. It was recognized that lack of management (principal) legitimacy strongly moderated expected control benefits over the physician (agents). A second, indirect, intervention focused on extending the informating concept to better fit the context of a clan of physicians. According to Ouchi (1979, p. 838), a clan forms when society demands a good or service that is difficult to control through explicit rules of bureaucracy or price mechanisms of markets. Under these conditions, members of a clan rely on control from
a deep level of common agreement between members on what constitutes proper behavior, and it requires a high level of commitment on the part of the each individual to those socially prescribed behaviors (Ouchi 1979, p. 838).
In clans, performance evaluation takes place through subtle reading of signals, ceremony, or ritual, that is possible among clan members "but cannot be translated into explicit, verifiable measures" (Ouchi 1980, p. 137). The second intervention, extending the informating concept to the clan, was recognized as a success, resulting in cost reduction and improvement in clinical outcomes. As was witnessed in this study, a clear advantage of action research is the opportunity to improve the system and its implementation as lessons are learned (Kaplan and Maxwell 1994).
The remainder of the paper is organized as follows: After the introduction of the problem situation in which the hospital faces financial challenges to reduce costs and a description of the action research methodology, Intervention 1 is introduced, where the hospital's management attempts to employ a decision support system (DSS) to directly informate the hospital's physicians. Intervention 1's learning highlights shortcomings in the principal-agent relationship implicit in direct attempts to informate when the principal lacks legitimacy. A second initiative, Intervention 2, is then undertaken by the management to mitigate its perceived lack of legitimacy by improving the recognized legitimacy of the information itself, by employing boundary spanning messengers, customizing user interfaces, and facilitating clan member discussion. This second intervention leads to the clan members' use of the provided performance benchmarking information in the physician group's concertive control processes and produces desired clinical practice changes. Finally, contributions are discussed followed by implications for research and practice.
The Problem Situation
St. John's Health System (SJHS) is an acute care community hospital in the Midwest region of the United States. During the 1980s, the costs of treating patients at SJHS continued to rise as the reimbursement for services changed from fee-for-service to capitation in which the insurance companies placed a cap on the amount of reimbursement. With the insurance companies shifting the risk of large cost increases on to the hospitals, SJHS felt threatened that its financial standing was at risk.
To exercise better control over its financial operations, SJHS needed to understand the service cost drivers. (3) Given that physicians drive as much as 80 percent of hospital costs (Chilingerian and Sherman 1990) and determine the quality of patient outcomes, the examination of physician practice patterns was a logical place to begin. Like many hospitals in the early 1990s (see Bloomfield and Coombs 1992; Covaleski et al. 1993), SJHS desired to build and implement information systems that better tracked physician-driven cost and quality outcomes to address these increasing institutional pressures. However, information systems in place at SJHS were not designed to track physician performance, nor was there a benchmark to compare such performance. Pressures from escalating costs grew to such a point that in 1991 the chief executive officer (CEO) summoned his senior administrative staff and top IT professionals to devise a solution. This action started a 10-year journey at SJHS to enact physician performance monitoring and behavioral change, which, over the course of this project, involved adjustments in theoretical understanding of IT's informating role and implementation schemes, as well as fundamental adjustments in the relationship between the hospital and its affiliated physicians. This paper chronicles this action research project, its failures and successes, its learning, and its theoretical contributions to better understanding the role of information systems in controlling professional agents when a principal lacks legitimacy.
Research Approach
Action research is an interventionist approach to the acquisition of knowledge that has its foundation in the post-positivist tradition (Baskerville and Wood-Harper 1996). Action research assumes that
a complex social process is best studied by introducing changes into that process and observing their effects. The theory underlying the interventions is validated by the extent to which these changes successfully solve problems in the setting (Baskerville and Stage 1996, p. 492).
Given that over time actors adapt and modify themselves, the technology and the evaluation process (Kaplan and Shaw 2002; May et al. 2000), action research has emerged as an approach for understanding hospital-physician interactions. Action research is also signified by the evaluation and control criteria for the project as well as the manner in which the researcher(s) interact and respond to the expectations of the clients. The evaluation criterion for this study incorporate action research characteristics outlined by Baskerville (1999a) and Susman and Evered (1978). Appendix A demonstrates how this study met these criteria.
[FIGURE 1 OMITTED]
This study's interventions follow the action research cycle (ARC) steps proposed by Susman and Evered (1978): (1) diagnosing, (2) action planning, (3) action taking, (4) evaluating, and (5) specifying learning. Figure 1 highlights the ARC steps for the two interventions. Table 1 provides a timeline for the study's interventions. For example, in Intervention 1, escalating costs were diagnosed as the problem and an action was planned. Consistent with the transparency tenets proposed by Zuboff's informating as well as the agency theory, an intervention was taken to directly introduce a DSS to the physicians so that their behaviors and outcomes were more transparent to the hospital management. The management hoped that such transparency could more directly influence physicians to cut costs. Evaluation of Intervention 1 indicated that while the DSS did make physicians' behaviors and outcomes more transparent, few physicians modified their practice behaviors. Intervention 1's learning led to Intervention 2, in which the action researchers diagnosed a need to extend the concept of informating into the context of a clan. Management undertook actions to indirectly introduce performance information into the clan's own concertive control process. For example, SJHS engaged a physician leader as a boundary spanner to help introduce a new physician profiling system (PPS). In essence conceding their own lack of management legitimacy, SJHS sought to elevate the perceived legitimacy of information provided by them in the eyes of the physicians. The evaluation and learning via the steps of the ARC process in the two interventions forms the basis for discussion in the next two sections.
Intervention 1
Intervention 1: Diagnosing
The first step in the Susman and Evered (1978) action research cycle--diagnosing--involves identifying or defining the problem. Typical of most acute care community hospitals, SJHS, as the principal, assigns the responsibility to physicians, as their agents, to dispense care to the patients. Physicians conduct clinical procedures, prescribe, and direct the nursing and ancillary clinical staff (radiology, pharmacy, laboratory, etc.) concerning patient care. In this way, cost is largely driven by the actions of physicians. SJHS delegates work to the physicians based upon their expertise by contracting clinical and surgical services with physicians and/or physician groups. These contracts are typically structured as hospital privileges, meaning that a physician has the right to practice medicine within a hospital based on the norms and rules set forth by a peer-regulated committee within a physician's practice area. Within this framework of hospital privileges, the physician is expected to offer quality patient care based on his/her professional norms of practice. For their services, the physicians either bill directly or get paid by the hospital based upon preset fee schedules; SJHS in turn bills for the total amount of cost incurred to the patients' insurance company. In either case, the dilemma faced by the hospital managers is that while they are bound by outside institutional forces to contain hospital costs, they do not have strong direct control over the primary agent (physician) driving cost and quality of patient care.
In 1991, the chief financial officer (CFO) of SJHS complained that the hospital was being reimbursed for its services through a predetermined formula instead of the costs incurred in treating patients. Furthermore, insurers were factoring the quality outcomes (length-of-stay, complications, and patient satisfaction) into the amount of reimbursement when renewing contracts. Thus, the hospital bore the risk of patient care costs and, since physicians' compensation was based upon preset fee schedules outside the control of the hospital, the CFO had little ability to alter compensation schemes to push the risk to the physicians. Also, accounting data showed a significant variation in hospital costs for the same procedures performed by different physicians within the same specialty. The CFO believed that an initiative that motivated physicians to reduce such variation toward lower cost and higher quality of procedures was needed. The CFO commented:
Its physicians' practice-driven costs that we need to get a handle on. We do not have comparative benchmarking information to make them aware of cost and quality improvement opportunities. In essence, we let them regulate themselves. (4)
During the 1980s, the CFO's office had made strides in exploiting IT to control costs incurred by nurses' and technicians' discretionary clinical behaviors. Specifically, the hospital implemented the productivity tracking system (PTS). The PTS was developed to track and standardize nursing and technical staff activities in providing care and had been very successful in calculating procedure-level costs. The implementation of PTS resulted in a significant reduction in variance among clinical activities. The PTS was a component of a larger clinical quality control information system that monitored clinical test orders and online results reporting. This system enabled computer-based monitoring of nursing and technical staff behaviors and, through benchmarking, training, and alteration in compensation schemes, resulted in cost savings. However, relative to the costs driven by the physicians, the overall cost savings resulting from PTS represented a relatively small portion of the hospital's total clinical costs. The CFO voiced his frustration to the director of continuous quality improvement and IT managers, including the manager of decision support systems in commenting:
If we could build a performance monitoring system similar to the PTS detailing [physicians'] costs and outcomes, and provide this information to the physician [practice area] committee, we might be able to set the right combination of pressure and cooperation to get them to make changes.
Similar to SJHS' experience with its PTS, other hospitals facing such problem situations have demonstrated management's enthusiasm to use performance monitoring information systems to contain rising health care costs. For example, Sia et al. (2002) report an implementation of an enterprise requirements planning system that gave a hospital "panoptic visibility," making nurses' and technicians' behaviors more transparent and strengthening processes of self, as well as, direct managerial control. In the case of performance monitoring of physicians, Bloomfield and Coombs (1992, p. 479) state hospital administrators pursuing a "resource management focus are very sensitive to the fact that the system will be, and is, seen as a new form of control over doctors."
In addition to defining the problem situation (as stated above) the diagnosis step of action research involves a search for theoretical assumptions to understand an organization's phenomenon or behavior (Baskerville 1999b, p. 15). Estab-lished theories can place the problem situation into a structure and guide the researcher to a successful conclusion of the project. While many bodies of knowledge, such...
NOTE: All illustrations and photos
have been removed from this article.
More articles from MIS Quarterly
Dialogical action research at omega corporation (1).(Special Issue), September 01, 2004
Small business growth and internal transparency: the role of informati..., September 01, 2004
Design principles for competence management systems: a synthesis of an..., September 01, 2004
Managing risk in software process improvement: an action research appr..., September 01, 2004
Looking for additional articles?
Search our database of over 3 million articles.
Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication
name or publication date.
About Goliath
Whether you're looking for sales prospects, competitive information, company
analysis or best practices in managing your organization,
Goliath can help you meet your business needs.
Our extensive business information databases empower business
professionals with both the breadth and depth of credible,
authoritative information they need to support their business
goals. Whether it be strategic planning, sales prospecting,
company research or defining management best practices -
Goliath is your leading source for accurate information.
|
.(special%20issue)%20%20&position=1)