Skip to main content
Back to How We Help Clients

Understanding patient preferences to boost performance of disease-management programs

Heath system sees 900 percent jump in diabetes education participation after redesigning the menu of interventions.


A European public health system struggled to determine why fewer than 10 percent of patients started, much less completed, a diabetes disease management education program after receiving a diagnosis from their doctor. The public health impact of unmanaged diabetes is clear, yet the reasons patients were not participating in support programs were not.

The diabetic population presented a challenging profile, as many patients had low household incomes, and the prevalence of the disease in these subgroups resulted in higher-than-average complications and deaths. The client asked McKinsey to help identify the causes of poor program attendance and develop new approaches that would increase participation to an aggressive goal of 70 percent.


From the outset, the project’s primary goals were to understand why participation was low and attrition so high, as well as when during the treatment and education cycle patients were dropping out of programs.

By interviewing patients and program staff members and analyzing program records, the McKinsey team identified several points along the patient treatment pathway where patients were likely to stop engaging in education programs. Some never sought out a practice nurse after getting a diagnosis, but many had at least one diabetes management session with a nurse before discontinuing the program. The largest drop-off—approximately 50 percent—occurred after introductory patient education sessions. Thus many diabetics did not take advantage of ongoing professional support or intervention from dieticians or specialists, even though they were aware of them.

In order to understand the causes of such steep attrition, the team worked closely with the client on two fronts. First, the team interviewed doctors and their staff members to understand what was happening in the doctors’ practices that could result in so few diabetics seeking treatment. Second, the team performed a detailed segmentation of the patient population itself using health records and information gleaned from patient interviews. Segmenting patients by disease state, treatment characteristics, mindsets, and barriers also helped the team evaluate the usefulness of various interventions.

Both efforts resulted in some valuable insights:

  • Doctors lacked incentives and demonstrated a mixed attitude toward patient education. They operated without standardized systems for informing patients about the support available, which ultimately limited referrals to about 6 percent of relevant patients.
  • Within the patient population, five broad groupings could be identified, categorized by different rational, emotional, and medical barriers, as well as different levels of willingness and ability to change their behavior.

The team then conducted additional analysis to evaluate the health service’s portfolio of educational interventions using four criteria:

  • patient enthusiasm for the intervention
  • ease of access
  • breadth of applicability to the diabetic population
  • cost per patient to administer

The patient segmentation and intervention evaluation work identified the forms of support that patients preferred. It also revealed that one primary cause of attrition was simply that many interventions (for example, a 2-day workshop) were inconvenient to attend.


With these insights, the team was able redesign the education programs so that more patients could participate more easily. Focusing on the top seven interventions (the original menu had more than a dozen), the rationalized portfolio of interventions was more cost effective to deliver and drove participation in education and disease-management programs to all-time highs.

Within 6 months of putting new processes and interventions in place, the client achieved dramatic results:

  • The number of patients completing an education course increased ninefold, leaping from 710 to 6,640, representing a 60 percent completion rate.
  • The number of total participants attending events approached 10,000.
  • Patient satisfaction increased significantly.

Health system managers continue to make adjustments to specific interventions to engage hard-to-reach patient groups, based on insights about patient preferences and known barriers to program participation.