Direct-to-Consumer Advertising: Public Perceptions of Its Effects on Health Behaviors, Health Care, and the Doctor-Patient Relationship

Development of Testable Hypotheses

We undertook a literature review to identify published claims about the effects of DTCA and formulate testable hypotheses.15

Sample

The survey was conducted between March 2000 and March 2001 on a household probability sample from the 48 contiguous states by using random-digit dialing to select households and a random selection method to select one respondent aged 18 or over for interview in each sample household. Verbal informed consent was obtained from all respondents before they began the interview. Eligibility was limited to English and Spanish speakers without cognitive or physical impairments that prevented completion of the interview. A small financial incentive was offered for completion of the interview. At least 15 attempts were made to call nonresponders, and where telephone numbers of nonrespondents could be matched with an address, up to 2 letters were sent to encourage response. The average administration time was 20 minutes. All interviews were conducted by trained interviewers using computer-assisted telephone interviewing (CATI).

An oversample of persons in poor health was achieved by screening a random subsample of households and including only those respondents who described their health as fair or poor (rather than excellent, very good, or good); had a disability or handicap that prevented them from participating fully in school, work, housework, or other activities; or had been hospitalized within the past 12 months, for reasons other than a normal delivery of a child. A total of 2720 interviews were completed in the nonoversample portion of the study, representing a completion rate of 72% and a response rate of 54%, and 489 interviews were completed in the oversample (completion rate, 95%; response rate, 51%).

Data were weighted to adjust for unequal probability of selection (including the oversample of persons in poor health). To adjust for survey nonresponse, stratification weights were developed using the March 2000 Current Population Survey from the US Census Bureau as standard. The poststratification weights were based on gender within age within race, as well as education, health insurance status (insured versus uninsured), and household size.

Interview and Data Collection

The interview was developed after literature review and focus group participation and piloted to ensure all questions were easily understood and contained no ambiguities. In the preamble, it was described as a survey on health issues, in particular how Americans make decisions about health care and their feelings about the health care available to them. No mention of DTCA was made in characterizing the survey. Initial questions inquired into respondents’ use of different sources of health information, approaches to health information, and relationships with their physicians. Respondents were then asked about their experience of searching the Internet for health information, because this was the primary focus of the project. Once questions about the Internet were completed, respondents were asked for their experiences of DTCA, whether they had encountered an advertisement that was personally relevant, and their response to that advertisement. Respondents who had sequentially seen an advertisement for a prescription medication in the past 12 months, perceived it as personally relevant, and discussed the information in this advertisement with their physician were asked about the last time they had done this. To avoid overloading respondents, those who had already answered questions about taking health information from the Internet to a visit with their doctor (n = 82) were not asked about the last occasion on which they had discussed information in an advertisement with their doctor.

Demographic and socioeconomic data, including age, self-defined ethnic origin, educational achievement, household income per annum, health insurance status, and current health status were collected from all respondents. Respondents were defined as being proactive about health care information if they stated that they went out of their way to look for information on health topics of personal relevance, rather than simply reading it if they came across it or not reading any such information.

Analysis

Dependent variables organized by hypothesis are summarized in Table 1. These variables include the kinds of requests made to physicians based on DTCA, the responses received to such requests, characteristics of the doctor-patient exchange over such requests, and the outcomes from the exchange. In addition, we included the doctor’s acting challenged as an important outcome variable, because it had been established as such in our previous work.16 We postulated that the doctor’s acting challenged could indicate either that the doctor felt challenged by interacting with an activated, knowledgeable patient or that the patient was failing to acknowledge the doctor’s professional expertise. Independent variables were categorized to maximize opportunities to determine the effect of low SES.

Univariate relationships between independent variables and the dependent variables were calculated using the χ² statistic or Fisher exact test as appropriate. Multivariate relationships were analyzed using the same methods of stepwise multiple logistic regression and adjustment for weighting described in the previous article.15

Characteristics of Respondents (n = 3209)

The characteristics of the respondents before and after weighting are presented in Table 2, which allows for an assessment of the demographic representativeness of our sample. The small differences between unweighted and weighted data suggest that the random-digit dialing method succeeded in generating a sample similar in profile to that of the US population.

Among the 308 respondents who had discussed information from DTCA with their physician, 82 were not asked about this because they had already responded to questions about taking information from the Internet to a visit (Figure 1). The 226 who did provide information were more likely to be unemployed (33% vs 20%, P = .036); less likely to be in good health (24% vs 43%, P = .011); and more likely to have a regular doctor (95% vs 82%, P = .008) than the 82 who did not provide information.

• Download figure
• Open in new tab

Figure 1.

Flow chart of respondents.

Overall Perceptions of DTCA

Most respondents were fairly positive about the recent increase in drug advertisements; 7% [95% confidence interval (CI) 6 to 8%] thought it was a very good thing and 40% (95% CI, 38 to 43%) thought it was a good thing. Thirty-four percent (95% CI, 31 to 36%) were neutral, and only 19% (95% CI, 18 to 21%) thought it was either very bad or bad. There was no association between socioeconomic status and overall opinion about DTCA. Table 3 presents the proportions of respondents who had seen a DTCA in the preceding 12 months who agreed with various statements about potential benefits and harms of DTCA.

Response to DTCA

Eighty-three percent (95% CI, 81 to 85%) of all respondents had seen an advertisement for a prescription medication in the preceding 12 months. The most common source of DTCA was television (94%), followed by newspapers or journals (62%) and the radio (22%). In the preceding 12 months, 20% (95% CI, 19 to 22%; n = 649) of respondents had seen a drug advertisement relevant to their health, and 10% (95% CI, 8 to 11%; n = 308) had discussed information in a drug advertisement with their doctors (Figure 1). Seventy-seven percent (n = 171) of these conversations were with a primary care physician.

Effects of DTCA

Table 1 summarizes our research questions, outcome variables, the number of respondents who stated they had experienced each outcome, and the significant associations with each outcome. Both in Table 1 and the text, the main outcomes are grouped into clinical benefits and harms, psychosocial benefits, effect on the doctor-patient relationship, and effect on health service utilization.

Clinical Benefits and Harms of DTCA

Psychosocial Benefits of DTCA

Effects of DTCA on Doctor-Patient Relationship

Effects of DTCA on Health Care System

Implications

We found significant positive and negative effects of DTCA on health behaviors, health service utilization, and the doctor-patient relationship. The impact was greatest on people of low socioeconomic status. This information is important to those seeking to maximize the benefits and minimize the harms of DTCA.

DTCA encourages members of the public, particularly those of low socioeconomic status, who are traditionally considered hard to reach with public health campaigns,11,12 to request preventive care and schedule a checkup. It encourages people to disclose health concerns to their doctor, and enhances some patients’ sense of confidence and control during a visit. From a policy or population health perspective, these undoubted benefits have to be weighed against the well-documented effect of DTCA on increasing health costs. In our previous article, we demonstrated that physicians experience DTCA as having an adverse effect on time efficiency and resulting in numerous, clinically inappropriate requests for tests, specialist referrals, and changes in medication. About half such requests are acquiesced to, with negative consequences on health care expenditure without commensurate health benefits. When considering the overall benefits and harms of DTCA, could the $2.5 billion spent on DTCA last year have achieved equivalent health benefits if the techniques of advertising were focused on interventions that are known to improve health outcomes?

Although only 5% of respondents who brought DTCA information to their doctors reported a negative impact on the doctor-patient relationship, 13% perceived that their doctor felt his/her authority had been challenged, and 21% reported sufficient dissatisfaction that they sought a second opinion, a change in doctor, or a change in health care plan. These data could represent the positive outcome of an informed health consumer or the negative outcome of information provided by a biased third party that undermines the consumer’s confidence in the health system. More research is needed to answer this question.

Methodological Issues

Several methodological issues limit the generalizability of our findings. First, this article reports on a secondary analysis of an extant data set, so not all hypotheses could be evaluated with the desired precision. In addition, lack of information from respondents’ physicians and/or medical records precludes direct verification of respondents’ perceptions. For example, we cannot determine whether requests made and filled were clinically indicated. Finally, respondents who discussed health information from the Internet with their physicians were not asked about the last time they took information from DTCA to a physician visit; hence, generalizability is restricted to the population that does not take information from the Internet to their doctor. However, these data are important because the 226 respondents who provided information about taking DTCA to a physician visit were of lower socioeconomic status than the 82 who did not. Other data from responses by all 3209 respondents indicate that DTCA has greatest effect on people of low socioeconomic status. To accommodate for this limitation, we conservatively assumed that the subsample that was not asked about discussing DTCA during a physician visit experienced no effects (ie, would not have sought an appointment or made a request). Thus, all our estimates of the population effects of DTCA are minimum estimates.