Patient adherence or compliance, defined as the extent to which patients take medications as prescribed by their healthcare providers, has been gaining substantial attention by pharmaceutical manufacturers in the past 5 to 10 years. While the World Health Organization (WHO) has reported in the last several years that only 50% of U.S. patients typically take their medications as prescribed, other reports indicate that the issue is more severe. According to a survey conducted last year by the National Community Pharmacists Association (NCPA), “Nearly three out of every four American consumers report not always taking their prescription medicine as directed.” Details of the report showed that:

• Half of those polled (49%) said they had forgotten to take a prescribed medicine;
  
  
• Nearly one-third (31%) had not filled a prescription they were given;
  
  
• Nearly three out of 10 (29%) had stopped taking a medicine before the supply ran out; and
  
  
• Almost one-quarter (24%) had taken less than the recommended dosage.

The National Council on Patient Information and Education issued a report last August, stating that poor patient adherence has resulted in an estimated loss of $177 billion annually in direct and indirect healthcare costs, including increases in hospitalizations and physician office visits resulting from non-adherence.

The severity of patient non-adherence differs widely depending on a multitude of factors including therapeutic category, patient demographics (age, income), product access, end-patient cost and stage in the therapy cycle. In order to develop effective patient adherence programs that increase awareness, companies must understand all of the factors that might impact non-adherence behavior including:

• Drug therapy and patient condition
  
  
• Past drug therapy/prescription history and trends
  
  
• Socioeconomic factors (region, culture, economic means)
  
  
• Patient-physician relationship dynamics
  
  
• Other decision influence channels (peer groups, news/media, etc.)

In the end, companies need to look at the complete patient profile, taking into consideration all of the factors that might influence non-complaint behavior (i.e., behavioral “triggers”).

Opportunity for Pharmaceutical Manufacturers

While improving patient adherence represents a significant opportunity for improving patient health outcomes, it also represents a substantial revenue opportunity for manufacturers. Brand teams seek solutions that help them more effectively target the right patient populations with personalized messaging while tracking and measuring the effectiveness of their patient adherence programs. To maximize the returns on their patient awareness programs, manufacturers require robust analytical solutions that can pinpoint the target patient population, develop the right messaging based on behavioral influences and measure the results of the programs (including ROI).

Predicting Patient Behavior: Targeting Models

There are a host of sophisticated analytical and modeling tools available to the marketing/brand teams today. The objective of most patient adherence programs is to identify behavioral triggers influenced by the market factors in order to find the most effective means to influence the patient. These programs typically involve increasing patient awareness of the benefits of drug adherence through a multitude of channels including media, pharmacies, physicians and direct (e.g., opt-in websites, mail). Since government regulations are strict on patient privacy and direct-to-patient communications, pharmaceutical and health services companies look for alternative channels to deliver the message. For example, an awareness program might include the involvement of physicians in order to identify the best patients for the “opt-in” awareness programs. Once the patient “opts in” to the awareness program, the companies can easily identify and provide personalized marketing to the targeted patient population. See Figure 1.

Figure 1

During the planning phase, brand managers need to define a targeting and segmentation strategy for the patient awareness program, which involves establishing the targeting criteria. One set of criteria used in the targeting and segmentation models includes demographic and regional factors. After analyzing adherence behavior for patients across regions and demographic groupings, marketers can establish patterns that help to identify targeted groups with higher potential adherence behavior.

One first step of the targeting model is to break patient populations into segments based on past adherence behavior. (Actual rate of refills % = (TRx – NRx) / Total prescribed (potential) rate of refills using anonymized patient-level prescription data.) For example, once the adherence rates within geographies are calculated, patient segments based on adherence “bands” (e.g., 80% - 100% = super- compliant) can be determined. Within these defined segments (super-complaint vs. semi-compliant), targeting models can allow marketers to further segment patient populations based on demographic factors and patient attributes. These factors and attributes will vary greatly depending on the drug therapy and behavioral “triggers.” They can include correlations between patient adherence rates and:

• Relationship to physicians
  
  
• Product access (e.g., plan/formulary changes)
  
  
• Socioeconomic factors (including income levels, population density, ethnicity)
  
  
• Attitudes towards health (e.g., gym memberships and health product purchases)
  
  
• Media and news channel influences (e.g., Internet, television, radio)

Most models developed for this purpose are predictive in nature; in other words, they will take a set of historical data points in order to attempt to predict (within a range of probability) patient adherence behavior. The output of the models will likely include territory and physician lists, geographical/geospatial (see Figure 2) and patient segmentation reporting. From the output, marketers can identify the target patient segments (down to the ZIP code level) and associated physicians and institutions (GPOs, hospitals, etc.) that need to be targeted. Based on a company’s territory assignment rules, these target lists can be translated to a list of the field sales representatives that need to be included as part of the targeted adherence program. The best models are updated continually throughout the year (e.g., monthly) in order that the criteria (inputs) forming the basis of the model may also be reassessed.

Figure 2

While the inputs used for a patient adherence program for a cancer drug, for example, will differ greatly from those used for a cholesterol-reducing drug, the models will likely share common reporting capabilities and need to be flexible enough to accommodate ongoing changes. The good news is that there is a high level of re-use once the models are built. Once built for a given therapeutic area, the models can be easily applied and leveraged to other programs for similar target and segmentation analysis.

The analytical software tools available in the market today have, for the most part, few limitations regarding the modeling capabilities necessary for building effective adherence models. However, not all brand teams follow a “best practices” approach to modeling. Before finalizing the targeting and segmentation schema for the patient adherence program, marketers should:

1. Assess past adherence behavior by geo-demographic factors.
   
   
2. Define complete patient profiles for low, medium and high adherence rates, including the behavioral “triggers” that induce adherence for the different segments.
   
   
3. Establish the criteria used to define targeted patient segments (inputs to the models).
   
   
4. Run predictive-based models that produce the patient segments (e.g., by geographies).
   
   
5. Establish the metrics (e.g., prescription market share) to be used for measuring program effectiveness.
   
   
6. Test the output of these models with smaller “pilots.” For example, you might want to redefine the targeting/call plans for a specific region or subset of the sales force to compare the results against your current targeting plans.

Sources:

Med Ad News Magazine

Centers for Medicare and Medicaid Services

SOURCE: Patient Adherence Programs: Components of Effective Targeting Models