Practice Pattern Analysis: A Tool for Continuous Improvement of Patient Care Quality.

Practice pattern analysis (PPA) is defined in this report as a method of aggregating data by practitioner, diagnosis, DRG, or other defined category to show patterns of care and/or variations in treatment. It is suggested that PPA is a tool that can assist physicians and other practitioners in acquiring meaningful information to support optimal clinical decision making and other evaluations. PPA, if used as a collaborative effort in hospital settings, permits meaningful information to be given to clinical decision makers. It is hoped that such feedback will enable practitioners to assess their clinical performance in comparison with their peer group. This potential is more likely to be realized if such data are provided in a multiphased, education-based approach, preferably with physician and clinical facilitators.

It is our contention that such information, encompassing small area analysis of a community hospital and its service area, will be perceived as being more relevant to the clinical practice under analysis. The database can be expanded when regional demographics and patient and provider characteristics are similar. Furthermore, it is imperative that PPA of individual practitioners encompass adequate numbers of patients to be statistically valid. Additionally, severity-of-illness adjustments must consider what individual patients bring to the clinical encounter. Other modifiers may recognize relevant physician and institutional characteristics for consideration in such analyses of performance. The monograph suggests that several key factors can contribute to acceptance of PPA by a hospital's medical staff. In our experience, these include:

* A shift in focus from one of reviewing not only outcomes but also processes of care. We have modified our professional review form to permit peer reviewers to identify opportunities for improvement (OFIs) that merit process analysis. We feel this allows the clinician to appreciate the change in philosophy of review from identifying practitioner outliers to improving patient care processes. It is our observation that resistance to review is reduced when this shift in focus is conveyed.

* Feedback of clinical and resource data to physicians in a confidential, professional manner encourages belief in the motivation for such activities to be quality improvement.

* Presentation of PPA in an understandable format, with supporting graphic materials, facilitates receipt of the information. Hospitalwide analysis is presented to the relevant practitioner group, while specialty group data are provided only at clinical departmental meetings, with individual practitioners listed but not identifiable. Practitioner-specific information is provided, as available, for the practitioner's perusal. It is otherwise available only in a trended format to the department chief at the time of reappraisal. The vice president for medical affairs and data analysis and quality improvement staff are available to respond to questions or concerns about these reports.

* A group of physicians serves in an advisory role to the data development department and the vice president for medical affairs in order to study techniques and prioritization of clinical study topics. Furthermore, clinical departments are asked to finalize the clinical conditions selected for study and the time frame for analysis and reporting.

In clinical departments, the use of comparative data and trend analysis is an objective discussion of appropriate or optimal clinical practice. Such data may be viewed along with clinical practice guidelines and consensus statements as enhancing clinical decision-making.

Many feel such collaborative uses of data to improve the appropriateness and the efficacy of clinical practice will have a positive impact on the bottom line of health care costs. It is hoped the reduction of disquality will enhance value while minimizing risks. As more sophisticated outcome measurements become available, the feedback process can be validated.

Economic credentialing has been promoted through the use of financial data in tracking utilization. If such data are used to develop criteria for medical staff reappointment, hospital-physician relationships will be strained. It is recommended that education-based PPA be incrementally introduced as the basis to encourage informed clinical judgments that are consistent with patients' expectations and the hospital's mission and resources. While some hospitals and lawyers say economic credentialing is a sound business practice, a consensus seems to be building that the use of economic criteria, unrelated to quality of care or professional competency, will be opposed. The credentialing process has always been the keystone of patient protection mechanisms of a hospital. It is imperative not to contaminate this endeavor with injudicious use of financial information.

Our experience has been that physicians often do not realize that their practices vary in comparison with those of peers. It is surprising how quickly physicians will alter their clinical decision making when provided comparative data and understandable information. In our experience, physicians will respond in a positive manner. We have, however, determined that continued feedback is required when more than 12 months have passed without providing information on a specific disease or procedure. There was some documentable regression to less effective and cost-effective practice patterns beyond that time frame. It has been our intention to design an educational data-based program that shares data of new disease/procedure categories as well as revisiting previous entities. Furthermore, it is our intention to use PPA as a tool for continuous improvement of patient care quality. We have avoided the uses of financial data outside the context of quality management activities. It has been our observation that such data prompt interest in, and discussion of, appropriateness and cost-effective use of resources. A physician advisory council, with the advice and consent of clinical departments, permits physician input into defining the goals of PPA in the hospital. This promotes the concept of openness as the hospital facilitates collection of data about physicians' practices, whether it be clinical or financial. Establishing and maintaining trust in hospital-physician relations is critical to the successful process of PPA. As noted in this report, there is a range of quality management activities to which PPA can be logically linked. At Anderson Area Medical Center, we include the following:

* Present clinical monitoring and evaluation activities, including analysis of possible nonphysician and hospital systems for variation.

* Risk management activities, including assessment of potentially compensable events (PCEs), as detected by incident reports and occurrence screens.

* Utilization management activities, with consideration of appropriate and necessary services.

* Results of external review, such as JCAHO, PRO, etc. In the latter instance, analysis of severity-adjusted outcomes by peer groups affords individual practitioners the opportunity to respond confidentially to quality and appropriateness concerns.

* Physician credentialing and privileging remains the cornerstone of patient protection activities within the hospital. PPA, as part of the individual physician's clinical performance profile, facilitates objective, data-based decisions.

* At our hospital, the vice president for medical affairs also serves as director of medical education. PPA data are recognized as a valid mechanism to define a meaningful focus for future educational offerings to physicians and other health professionals.

* Summary reports and comparative quality indicators will satisfy the governing body that quality and financial management factors are being analyzed and addressed. The board member who chairs the board quality committee sits on the quality coordinating council to review clinical quality activities.

Each hospital must go about the process of defining and meeting the data needs for PPA. This analysis must proceed within the context of the quality improvement philosophy. The AHA report suggests the following caveats (I have added some of my own!):

* The sources of the data must be understood, avoiding proprietary "black boxes."

* The data must be accurate and timely so that resulting analysis provides information that reflects current practice.

* The definitions that are used in choosing, compiling, and comparing data sets must be clearly understood and applied so that valid comparisons are made.

* Except for use by individual practitioners for educational purposes, the data format should protect the confidentiality of practitioners.

* Data analysis should focus on patterns of care instead of individual occurrences.

* PPA should operate under the ideal of continuous quality improvement (CQI), as practitioners strive to improve their clinical decision making and relevant hospital processes.

* Clinical department chiefs and other physician leaders should assume responsibility for promoting use of the information resulting from PPAs. Specialty group discussions should occur, with the goal of stimulating interest in the data and acting on the conclusions. When appropriate, the chief may initiate one-on-one dialogue.

* The prevailing theme should be that PPA will remain an objective, educational effort that has as its goal the improvement of patient care.

The advisory committee should set the goals and the uses of PPA. The hospital, and its medical staff, should establish which services and procedures need PPA. The services and procedures may be:

* High risk.

* High volume.

* Suspected significant variability.

* Issues identified by monitoring and evaluation and other quality assessment activities.

* Liability concerns.

* Results of published clinical studies or external outcome data.

* High-cost procedures or conditions.

* Procedures or conditions found to be associated with significant financial adjustments or variations in resource usage, not accounted for by acuity considerations.

* Clinical situations with unexpected outcomes-preventable deaths etc.

* When discrepancy is recognized between hospital practice and valid clinical benchmarks.

A word of caution! It would be unwise to initiate practitioner-level analysis until data collection and management systems can match expectations. Once useful data can be provided, it is also important that confidentiality be protected and that data be reported in a useful format. Institutional support is required to ensure that these specifications are met. The purchase of data management systems should be made with the concurrence of the medical staff. This report suggests that the process of PPA is a complex one that will require a long-term commitment. It will require not only the personal and professional energies of hospital executives and medical staff members but also the judicious allocation of hospital resources. Nevertheless, PPA is well worth the effort. In particular, according to the report, if it is performed with care and mutual trust, it can strengthen the critical partnership between the hospital and the medical staff. It can help integrate the two sometimes parallel but often adversarial processes of quality and cost management, and it can position the hospital to better respond not only to individual patient care needs but also to the larger demands of the external environment. Given the opportunity to achieve these benefits, progressive hospitals and medical staffs will begin to explore PPA with optimism and will eventually embrace it with enthusiasm and commitment.