US patients who fail to take their medications as prescribed exact a staggering economic and social cost.¹ A McKinsey study of people with hypertension (high blood pressure) suggests that a better understanding of the attitudes of patients could improve programs designed to increase their adherence to treatment regimens—an outcome that would save both lives and money. Creat-ing and implementing these programs will require the combined efforts of physicians, patients, pharmaceutical companies, payers, and other health care stakeholders.

Hypertension afflicts 65 million people in the United States, fully half of whom don't adhere to their drug therapies.² The usual interventions—for example, electronic reminders or easy-to-open packaging—tend to improve adherence only in the short term, largely because a one-size-fits-all approach fails to address the underlying causes of the way patients behave. To improve upon these programs, we surveyed 810 hypertensive patients and supplemented their responses with interviews and focus groups.³ We also asked patients about two other chronic asymptomatic medical conditions—type 2 diabetes and high cholesterol.

We explored five themes: the patients' level of involvement and perceived control over their health, their knowledge of hypertension and its treatment, their level of concern about the disease, their beliefs about the safety and efficacy of medications, and the quality of their interactions with physicians. Survey respondents indicated their level of agreement with statements such as, "I am very active in the management of my health" and "My high blood pressure is not very serious." In addition, the survey explored the patients' receptiveness to interventions that might influence their degree of adherence.⁴

When we analyzed the relationship between the attitudes and self-reported behavior of patients, six segments emerged (Exhibit 1, part 1). They ranged from proactive patients, who scored high on all five themes (Exhibit 1, part 2), to skeptical patients, who, for example, thought their condition was not serious, distrusted both physicians and medications, and adhered to their regimen as infrequently as 5 percent of the time. Likewise, we observed strong links between behavior and attitudes toward medication among the remaining segments: confident, concerned, confused, and resigned.

Comparable segmentation approaches that provide for targeted interventions or messages focused on the key barriers to specific activities have aided many marketing efforts (including antismoking campaigns). But the approach has not yet been widely used to improve the adherence of patients to their treatment regimen. Our study suggests that it should, since different types of patients have very different reasons for not taking medications: we found that information on their long-term safety might quell the fears of concerned patients, for example. By contrast, confident patients (half of whom mistakenly believe that they could safely discontinue therapy in the future) generally adhere to their treatments and might benefit from programs that reward consistent, long-term usage—for instance, loyalty-type programs, (with points for refilling prescriptions or that give patients free gym memberships). Such programs would support the confident patients' high level of interest in managing their own health.

Resigned patients indicated that following a healthy lifestyle is too much trouble for them, yet they were extremely concerned about the risk of a stroke. More than 80 percent of the patients in this segment would be more likely to take their pills when presented with the facts about the link between hypertension and strokes. In addition, only 7 percent of the resigned patients have a routine for taking their medications, so there is a need for simpler reminder devices. In fact, more than 50 percent of these patients (almost twice the average) expressed an interest in them (Exhibit 2).

Implementing tailored interventions targeted at the root causes of nonadherence could help patients across a wide range of conditions. In fact, we found that respondents with both high cholesterol and hypertension reported similar attitudes and behavior about taking their medications, suggesting that this segmentation and the resulting interventions would work across both conditions. However, the adherence of respondents with both hypertension and diabetes followed a different pattern for each condition—perhaps because the more intrusive treatments for diabetes (such as insulin injections) lead patients to perceive it as a more serious condition than hypertension. This finding suggests that as conditions manifest more symptoms, both the segments and interventions change. New research is required to develop the required insights into the attitudes and behavior of patients.

Ultimately, programs developed along these lines might involve patients, physicians, payers, employers, pharmaceutical companies, and even government agencies. Physicians, for instance, could identify patients by segment and then tailor interactions accordingly, whereas payers could use these insights to develop more targeted programs for employers. Pharmaceutical companies—which have marketing resources and expertise, as well as extensive relationships with physicians—are uniquely placed to take the lead in recruiting other parties.

Of course, collaborating on such a scale to design programs brings its own challenges—notably the need to keep information about patients confidential. But recent trends, including efforts by companies to make employees more responsible for health care choices⁵ and calls for physicians to be more accountable for their performance, may provide the impetus to form these coalitions. The stakes are too high to ignore. Indeed, the World Health Organization notes growing evidence that increasing the adherence of patients to their regimen could improve public health more than advances in specific medical treatments.