For the past decade, the healthcare industry has struggled to identify methods to significantly modify clinical practice. A recent review of typical educational programs confirms the inability of traditional continuing medical education to change clinical behavior.¹ Introducing evidence-based clinical guidelines also has failed to improve clinical care.^2,3

The need to identify reliable ways to improve clinical care has led several teams to prepare and test new types of interventions. Avorn and Solomon³ identified specific interventions that improved appropriate antibiotic use. These approaches included reminders at the point of care, academic detailing, and order entry programs. In 2001, Grol⁴ argued that the complexity of changing clinical practice behaviors requires more than a single intervention such as an educational program, financial incentive, or practice profile. To promote successful practice outcomes and adherence to guidelines, Grol proposed creating an integrated combination of self-reinforcing interventions such as evidence-based guidelines, professional education, assessment and accountability, patient empowerment, and total quality management. Bodenheimer, Wagner, and Grumbach^5,6 also recently endorsed a multiple intervention, multilevel model for improving chronic disease care.

The primary aim of this study was to demonstrate the ability of such a multifaceted intervention program to improve the evaluation and management (E&M) of acute sinusitis, especially in regards to appropriate antibiotic prescribing. Interventions included physician education, a locally developed acute sinusitis care pathway, feedback through a physician profiling system, a financial incentive for adherence to our care pathway, and patient education. The intervention was applied to a community-wide panel of more than 900 primary care physicians covering 420 000 health maintenance organization (HMO) members. The high penetrance of the HMO in the local market allowed us to examine and profile individual physicians on large numbers of cases without needing to pool data among multiple payers.

METHODS

Physicians

This project evolved from a collaboration between the Rochester Individual Practice Association, Inc (RIPA) and BlueCross BlueShield of the Rochester Area's HMO, Blue Choice. (BlueCross BlueShield of the Rochester Area has since been renamed Excellus BlueCross BlueShield of the Rochester Region.) The baseline measurement period was January 1, 1999, through October 31, 2000. The intervention period started November 1, 2000, and was measured through December 31, 2001. In 2000, RIPA was a communitywide panel of approximately 3000 practitioners and 900 primary care physicians (500 internists, 200 family practitioners, and 200 pediatricians) serving 420 000 HMO subscribers. Rochester Individual Practice Association, Inc, physicians were located in the 9-county region surrounding Rochester. The region includes urban, suburban, and rural communities.

Credentialing information was used to identify all internists, family physicians, and pediatricians. All credentialed physicians actively seeing patients were included in the program. Cases of acute sinusitis seen by nurse practitioners or physician assistants were assigned to their supervising physician.

Patients

Analysis included all HMO patients treated by the identified physicians or their nurse practitioners and physician assistants. The HMO provided an administrative data set with scrambled patient identification numbers. Because neither patient-specific information nor reviewed medical records were used, informed consent was not obtained.

Case Identification

Episode treatment group (ETG) methods were applied to claims data to identify cases, or "episodes," of acute sinusitis without sinus surgery and their related services (Episode Treatment Grouper, version 4.0; Symmetry Health Data Systems, Inc, Phoenix, Ariz; incorporated in Episode Profiler, version 4.6, CareEnhance Resource Management Software; McKesson Health Solutions, LLC, Cambridge, Mass). In the ETG model, an episode consists of a series of healthcare services related to a group of specific diagnoses for 1 patient. An episode of care may contain single or multiple services, including office visits, diagnostic tests, therapeutic interventions, emergency department visits, and prescribed medications. Episode treatment groups are structured so that each contains 1 disease entity (as is the case for acute sinusitis) or clinically similar diseases.

Analysis of each episode began by finding an anchoring E&M service. The anchoring E&M service could have an International Classification of Diseases, Ninth Revision (ICD-9-CM) code for acute sinusitis or for a more general illness such as a "viral upper respiratory tract infection." The software examined each subsequent claim by procedure per Current Procedural Terminology codes, by diagnosis per ICD-9 codes, or by medication generic code number and national drug code. It also searched backward 1 month for antibiotics prescribed before the initial patient encounter. The software then decided whether to add the claim to the current episode of care and, if so, whether the episode should remain in the current ETG or be assigned to another ETG. For example, if the episode began with a viral upper respiratory tract infection and then later had an E&M service for acute sinusitis, it was removed from the former ETG into the sinusitis ETG. If sinus surgery occurred subsequent to the anchoring visit, the episode was transferred to the ETG for sinusitis with surgery. An episode was considered completed when no additional services were billed for the ETG for a fixed period (the "clean" period).