Thursday, July 28, 2011

CKM and Positive Deviance

Positive deviance is a strategic approach to identifying top-performing individuals or groups, and disseminating their special knowledge to the remainder of the group. If you've read Malcolm Gladwell's book Outliers, you are familiar with the concept of positive deviants; positive deviance identifies these outliers for the basis of process change and improvement. Historically, positive deviance was used to improve conditions in the developing world, including malnutrition and peri-natal mortality. Recently, positive deviance has been applied to quality improvement in health care.

Positive deviance can play an important role in health care. The process not only identifies the positive outliers but also provides the framework for disseminating their knowledge. However, until recently, identifying individuals and groups with significantly better outcomes has been difficult in health care.

As organizations increasingly implement robust HIS systems, the systematic identification of positive deviants has become possible. From this starting point, the AURI cycle can be used to create actionable knowledge which can be disseminated and implemented.  This is why clinical knowledge management is an excellent platform for quality improvement.

To use positive deviance in the CKM process for a QI project, the "analyze" step of the AURI cycle is divided into three actions:

  1. Identify individuals or groups with significantly better performance in your outcome of interest
  2. Utilize quantitative techniques to determine differences in care delivery between positive deviants and others
  3. Utilize qualitative techniques to determine differences in care delivery between positive deviants and others
During the "understand" step of the AURI cycle, the differences in care delivery identified in the "analyze" step are examined to determine which are most significant. These significant differences are then incorporated in the "redesign" step. During the "implement" step, these differences in care delivery and their importance to the redesign are communicated to the group. The cycle continues by analyzing whether the remainder of the group has adopted the new care processes and if they have had the desired impact on outcomes. 

By using positive deviance in conjunction with the CKM process, quality improvement projects can demonstrate rapid improvement. 

Monday, July 25, 2011

CKM and Readmissions

The Journal of General Internal Medicine recently published my paper examining the effects of standardized discharge instructions on readmission. Readers may be surprised that standardizing our discharge instructions to meet consensus recommendations did not reduce readmissions.

However, in light of the necessity of providing actionable knowledge to improve decision making, these findings begin to make sense.

The standardization of the discharge instructions focused on insuring that a set of recommended components (ie: discharge medications, follow-up appointments, contact information, etc.) were always provided to the patient at the time of discharge. On the continuum from data to actionable knowledge, these components are information. Although patients can often synthesize these various information points into knowledge, good decision making is born of actionable knowledge. To move from information, to knowledge, to actionable knowledge, you first have to understand the question that discharge instructions are trying to answer: how should I take care of myself outside of the hospital so that I don't have to come back? 


In light of this question, actionable knowledge is based on an understanding of the patient's own health and disease processes as well as the actions necessary to maintain their health. The capacity to understand and perform these actions is highly variable among patients. Therefore, interventions to prevent readmission need to be customized to each individual patient rather than standardized.

Whether these customized interventions will be time efficient and cost effective has not yet been determined. However, in my opinion, discharge interventions that are not customized to the patient will continue to show lackluster results.