Think globally, act locally: an approach to implementation of clinical practice guidelines

In an environment of concern about the rising costs of medical care, the Vermont Program for Quality in Health Care (VPQHC; the program) was incorporated in 1988 as a nonprofit organization and in 1989 was made a peer review organization by the state legislature. The program acts a resource center for health care in Vermont, coordinating three functions: implementation and maintenance of a statewide database for healthcare quality; training for health care providers in continuous quality improvement (CQI) methods and support for their CQI projects; and focusing clinical study group work on specific diagnoses or procedures.

The program uses a seven-step process for implementing CQI: pick a process (modify a nationally developed guideline or develop a new guideline); select a team of people involved in doing the work; establish goals and key quality factors; document the clinical process; determine what and how to measure; measure and analyze data; and modify the process to improve. GUIDELINE: This article describes the implementation of a guideline from the American College of Obstetricians and Gynecologists (ACOG) on cesarean section (C-section). Except for a few specific contraindications, the ACOG paper states that it is reasonable to encourage vaginal birth after C-section. A related set of criteria from ACOG states that the benchmark for emergency C-section should be 30 minutes from the decision to proceed with an emergency C-section until the baby is born.

State C-section rates from 1985-1990 showed wide variation by hospital. Such wide variation is based as much on practice style as on sound science. C-section rates (primary and repeated) were measured, plotted, and shared with Vermont hospitals. Successful vaginal birth after C-section (VBAC) rates from the statewide data-base also were shared. Based on these data, one hospital, Hospital A, developed a plan to lower its repeated C-section rate and improve its VBAC rate. Hospital A collected and reviewed local data, which showed interesting variations. The time from decision to birth became the focus of the overall project; Hospital A designed its project to study events during the time from decision to birth.

Most recent data (1992-1993) from Hospital A shows improvement in three areas. First, 89% of patients having emergency C-sections met the goal of a 30-minute time frame from decision to incision. Second, the VBAC success rate for the same time period increased to 85.7% from 69% in the previous year, and from a mere 7% before the implementation of the project. Third, the percentage of total C-sections that were repeat C-sections fell to 36% from a high of 51%. In a follow-up evaluation, one-third of the C-section performed from October 1, 1993, to July 31, 1994 were performed because of patient refusal to attempt VBAC, suggesting that there is a communitywide culture that influences behavior. A major effort at patient education on VBAC is underway.

Think globally, act locally, might be the motto for the program. Implementing guidelines starts with obtaining national guidelines and literature but needs the use of local data to sharpen the focus on narrow areas to address. Specifically, it is unrealistic to tackle the entire problem at once. Success comes from finding specific opportunities for improvement.