HMI World

Welcome

Hospital honor roll ranks two Boston hospitals among the nation's best

Letter From The Editor

We don’t build each issue of HMI WORLD around a common theme, but sometimes those themes emerge on their own. The theme that jumps out in the November-December 2007 issue is outputs.

Around Harvard

Medication reconciliation helps prevent drug errors

Small clinical center uses simple paper-based system to reduce mistakes

The Institute of Medicine estimates that at least 1.5 million preventable adverse drug events (ADEs) occur every year in the United States—an average of one medication error for each hospitalized patient per day. Not only do these errors directly affect patient health, but they also cost more than $3.5 billion dollars a year. Eliminating preventable ADEs has become an important objective for health care organizations. However, the goal is a challenging one for several reasons. Medication orders frequently change as patients transition between home and inpatient facilities and between inpatient care units. Orders commonly are written by a large number of different care providers. In addition, processes to find and correct any discrepancies are often lacking.

Communication problems are a common cause of medication errors. Of the more than 350 medication errors resulting in death or major injury included in the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) sentinel event database, 63% were related to failed communication. Medication errors tend to occur at transition points during the continuum of care. A 1995 study by Bates and colleagues found that almost half of all preventable ADEs in the inpatient setting occurred during ordering, a process that happens at three key transition points: admission, transfer between inpatient care units, and discharge. A more recent study by Cornish and colleagues found that 54% of patients in a general medical unit experienced a medication error at admission, most commonly due to failure of the staff to include a regularly used medication.

Medication reconciliation is a process by which a patient’s current medication orders are compared with a list of previously taken medications. The process involves three basic steps:

Collecting an accurate medication history
Checking that medications and doses are appropriate
Documenting all medication changes in the orders and making sure these changes agree with other medication information

According to JCAHO, medication reconciliation can prevent drug interactions, omissions, duplications, and errors in dosing. Research has shown that the process can reduce medication errors. For example, a 2006 Canadian study by Vira and colleagues found that the use of medication reconciliation identified and prevented 75% of clinically important medication errors.

SOUNDING BOARD: Has your institution undertaken a quality improvement project recently? What were the goals? How was the project organized? What were the outcomes? What key lessons learned might benefit others attempting similar projects? We'd like to hear more about what health care organizations are doing to enhance quality and improve patient safety. Write to us!

Indeed, the process of medication reconciliation is no longer optional for many institutions. As of January 2006, the JCAHO National Patient Safety Goals mandate that accredited facilities must have protocols in place for documenting and reconciling medications during the continuum of care. The mandate has prompted many health care organizations to implement medication reconciliation programs. However, putting into practice a process that is new and involves system-wide changes can be challenging. It can be labor-intensive, time-consuming, and costly, especially if new electronic database systems are used.

Clinicians and administrators frequently encounter obstacles when implementing a medication reconciliation program. John Whittington, MD of the Institute for Healthcare Improvement notes that the most important of these challenges is confusion due to a lack of clearly articulated roles and responsibilities. Such ambiguity can lead to oversight errors or to turf battles. Other possible challenges include lack of staff engagement or “buy in,” staff concerns about the additional time necessary for the process, and resistance to change. Despite these challenges, some health care organizations have successfully implemented reconciliation programs. Contra Costa Regional Medical Center is an example of one such organization that was able to make medication reconciliation a “win-win” proposition for patients and staff, reducing errors while streamlining the work of clinical staff.

Contra Costa Regional Medical Center (CCRMC), located in Martinez, California, comprises a 166-bed hospital and eight federally funded health centers. After the organization decided to implement a medication reconciliation program, a multidisciplinary team—consisting of pharmacists, nurses, physicians, quality management staff, and a nurse program manager from clinical informatics—was charged with putting the process into practice. The team’s first step was to develop a plan based on input from four hospitals with advanced medication reconciliation programs already in place. A crucial early decision was the team’s choice to divide the intervention into manageable portions, thus avoiding simultaneous execution of a system-wide program. Instead, the team began its work by focusing on reconciliation at hospital admission within a single eight-bed telemetry unit.

The team first considered the current workflow: a resident physician or attending physician collected a list of home medications in the chart’s history and physical (H&P), a nurse collected a second medication list at his or her admission evaluation, and the physician wrote a third medication list as part of the admission orders—three different lists that were not systematically cross checked. Through small tests of change—testing a process, rapidly assessing it, making necessary changes, and immediately testing again—the team eventually developed a single medication list that replaced the previous three lists.

In the revised system, the physician captures the list within the orders; medications are no longer listed within the H&P or nursing admission note. On the order sheet, the physician must indicate for each drug whether it is to be discontinued, continued, or modified. The CCRMC process also captures a list of home medications, which is later used at discharge for planning and reconciliation. Once created, the list is faxed to the pharmacy, where a pharmacy technician enters the information into a database system and a clinical pharmacist reviews the list.

After rolling out the project to a 30-bed unit, news about the project and enthusiasm for it spread rapidly. In fact, resident physicians and nurses began using the process in a medical-surgical unit that had not yet received training. (The team asked the staff to temporarily stop using the process, to avoid uncontrolled spread without education.)

Once unit-by-unit roll out of medication reconciliation at admission was successfully underway, the team focused on reconciliation at transfer. After a similar series of small tests of change, the team developed a process in which the physician conducts a formal reconciliation when writing orders for the patient’s transfer. The orders must indicate whether to continue, discontinue, or modify every medication on the medication list.

Finally, the team focused on reconciliation at discharge. Using the small tests of change again, the team developed a process in which the discharging physician reviews the current hospitalization medications and the list of home medications captured at admission. The review process serves as a reminder to specifically inform patients which medications to restart upon discharge and which medications not to restart at home.

Full implementation of the medication reconciliation process at CCRMC took two years, from mid 2005 until mid 2007. The time investment has paid off in reduced medication discrepancies at key transfer points. The proportion of pre-admission medications that were unreconciled on admission decreased from 25% at project initiation to 4% after full adoption of the new process. The proportion of medications unreconciled at transfer dropped from 12% to 4%, and the proportion of patients with any hospital medication unreconciled at discharge was reduced from 36% to 2%.

According the Steven Tremain, MD, ABFP, FACPE, Director of System Redesign at CCRMC, the team’s success was dependent on several factors. First, beginning with a small unit with a regular stable of staff proved to be ideal for pilot testing. Second, the organization combined both “push” and “pull” strategies to engage staff in the intervention. The JCAHO mandate for reconciliation served as a push. The team ensured that several “pull” strategies were in place by involving front-line staff in every step of the project—not only did staff members “own” the system that they helped developed, but also the system was better suited for their particular needs because of their input. The clinical staff wanted to adopt the process, because the intervention eliminated redundant work, making their jobs easier, and it reduced the guesswork errors that had been associated with medication orders.

Tremain believes that a key element of the team’s success was recognizing the importance of earning staff trust. The team found it absolutely essential to inform staff members of three key pieces of information before beginning any test of change: the specific element of the intervention being tested; the test start date; and the test end date. “You must let the front-line staff know that a test of change is not going to continue indefinitely. People are much less resistant to trying something for a defined period of time.”

For additional information and resources, please click here.

--Written by Diane Shannon, MD, MPH for Harvard Medical International