Utilization of adequate provision in prescription drug broadcast ads among low- and non-Internet users

Abstract

To fulfill a key regulation pertaining to prescription drug broadcast advertising, a common practice is to present the major risks along with “adequate provision” referencing several sources where audiences can obtain the full product labeling. In recent years, questions have arisen about the unique value of the various sources of adequate provision and even whether sponsor webpages alone may sufficiently convey product labeling information. Cognizant of the sizable offline population, the present research investigates questions of access, ability, likelihood, willingness, and preference among a nationally representative sample of low- and non-Internet users who may wish to access the product labeling.

Introduction

Prescription drug advertising regulations require that broadcast (e.g., television) advertisements containing product claims also present the product’s major risks via a major risk statement (21 CFR 202.1(e)(1)). The regulations also require that broadcast advertisements contain a brief summary of all necessary information related to side effects and contraindications, or alternatively, that (in addition to the major risk statement) “adequate provision” be made for dissemination of the approved or permitted package labeling in connection with the broadcast. The common practice for fulfilling these regulatory requirements is to include the major risk statement together with adequate provision.

U.S. Food and Drug Administration (FDA) guidance (FDA, 1999) lays out several ways through which approved product labeling can be disseminated, such as through a toll-free telephone number; print advertisements appearing concurrently in publications that reach the exposed audience or other print media like product brochures or information available at pharmacies or health care provider offices; and through the Internet. The goal of including these elements, plus a reference to a health care provider as an information source, is to ensure that diverse audiences, regardless of their abilities, preferences, or privacy concerns, can conveniently access the information.

In recent years, questions have arisen from industry leaders regarding the necessity of requiring all the suggested components of adequate provision. In our increasingly digital age, it has been suggested that most consumers may be able to access the product labeling online, from the convenience and privacy of their homes, or in cases where the Internet is not available, through public libraries and the help of family and friends. Although these notions are attractive conceptually, and although the Internet has been reported as the most-used source for this information (Aikin et al., in press), a segment of the population continues to report that the Internet is inaccessible to them (Anderson et al., 2019).

As of June 2019, 10% of people in the United states did not use the Internet (Pew Research Center, 2019a) which equates to approximately 32.8 million people (Census Bureau, 2020). Further taking into consideration those who use the Internet only infrequently—for which estimates are not currently available—a substantial number of individuals in the United States are not regularly accessing the Internet. Moreover, various demographics are known to be less likely to use the Internet. For example, approximately 27% of those older than 65 do not use the Internet (Pew Research Center, 2019a). Older adults are of special relevance because they also take more prescription drugs than younger people (Kantor et al., 2015), watch more television (Depp et al., 2010), and are more likely to experience adverse drug reactions (Lavan & Gallagher, 2016). They are also more likely to miss important information when viewing prescription drug broadcast ads (O’Donoghue et al., 2019) and so may benefit from the opportunity to review information made available via adequate provision at their own pace.

The digital divide, or the disparities that arise between those with and without access to the Internet, is especially concerning for health outcomes (Kreps, 2005). Access to health information is critical to guiding healthcare decisions and a lack of access to the Internet is known to impact populations that already face health inequities (Viswanath & Ackerson, 2011). For example, individuals with lower socioeconomic status are less likely to both access and have trust in health information on the Internet (Viswanath et al., 2007). Additionally, confidence in using the Internet, ability to navigate complex information online, and willingness to use the Internet as a source all influence how individuals might be able to utilize health information on the Internet (Cline & Haynes, 2001; Dutta-Bergman, 2004). Therefore, it is important to consider access, ability, and willingness when assessing how a population may be able to use health information found exclusively online.

Lack of Internet access and ability among various demographic groups prompts several questions relevant to the issue of adequate provision, which we seek to address in the present research. First, we ask Research Question 1, or RQ1: Are low- and non-Internet users able and willing to access the various sources commonly used together to fulfill adequate provision requirements? A reasonable assumption is that these individuals require an offline method, like print materials or a toll-free number. Today, most individuals (~96%) in the United States have access to a mobile telephone (Pew Research Center 2019b) and for many, dialing a toll-free number may provide greater convenience than tracking down print materials. Consequently, print materials such as print ads and brochures disseminated independently from a toll-free number may no longer offer unique value. Thus, we also ask RQ2: Are the print component and toll-free number both important, or are low- and non-Internet users able and willing to access the product labeling in just one of these ways, or in other ways? These questions assess how these individuals might prefer to gain access to the product labeling with special consideration for their low- or non-Internet use.

We also ask RQ3: Are low- and non-Internet users able and willing to seek out the product labeling online if they have a specific information-seeking goal and if the information is only available online? Many offline individuals may be offline by choice. That is, they may be capable of accessing the Internet but choose not to. At the same time, some offline individuals may not currently know how to use the Internet but are willing to learn. Likewise, some offline individuals living in areas without Internet connectivity may be willing to travel to access such connectivity. This research question investigates whether the offline population can be pulled online in a scenario in which print materials and toll-free numbers are not offered by drug sponsors.

The FDA guidance (FDA, 1999) also acknowledged that consumers may desire privacy when seeking product information. Consumers frequently express concern about privacy related to their health information (Bansal, Zahedi, & Gefen, 2010) and FDA’s guidance recognizes that these concerns may extend to information seeking about prescription drugs. As noted earlier, it has been suggested that offline individuals can access the Internet with help from their family and friends or via public libraries. Although this approach may provide the offline population with Internet access, it may not offer the desired privacy. Consequently, print materials and/or the toll-free number may provide significant value to privacy-conscious individuals. In response, we ask RQ4: What are low- and non-Internet users’ concerns about privacy related to accessing product labeling, if any?

Based on the issues and literature summarized above, we conducted a nationally representative survey of the offline population to investigate their perspectives on accessing product labeling via the various components of adequate provision. Our assessment examined this population’s ability and willingness to use the Internet, toll-free numbers, and print magazine ads as sources of adequate provision, and their preferences for using these sources to obtain product labeling information when they have privacy concerns.

Method

Sample and Procedure

The sample consisted of 1,486 adults collected from an omnibus survey, a nationally representative survey of random-digit dialed (RDD) landline telephone households and randomly generated cell phone numbers. The sampling approach involved using an RDD sample of pre-identified households without Internet access. The study was restricted to adults aged 18 years or older. To qualify, participants had to indicate that they had browsed the Internet two or fewer times in the last month, watched television at least occasionally, and currently take or had ever taken at least one prescription medication on a regular basis. The survey was administered in both English and Spanish through a computer-aided technology interviewing (CATI) system.

Study Measures

Participants answered demographic questions about their age, race, ethnicity, education, and household income during the screening process. Then, the survey measured accessibility and preferences for accessing product labeling through the company’s web page for the drug on the Internet, a toll-free number maintained by the drug company, and a print ad in a magazine. For access, ability, and willingness questions, the survey distinguished between using the Internet with and without help from another person, as well as at home or outside of the home.

Internet, Phone, and Print Magazine Ad Accessibility.

We measured general access to Internet, phone, and ads appearing in print magazines by asking participants if each was available to them (“Would you say each of the following is available or not available: [insert method]?”; “not available”/“available”).

Ability to Use Internet, Phone, and Print Magazine Ads.

We measured general ability to use Internet, phone, and ads appearing in print magazines by asking participants a closed-ended question for each: “Would you say you have no ability, some ability, or complete ability to use (insert method)?” (“no ability”/“some ability”/“complete ability”).

Likelihood of Accessing Product Labeling Using Adequate Provision Delivery Methods.

We asked participants to imagine they wanted to obtain additional information about a prescription drug they saw on television before seeing their healthcare provider. We assessed their likelihood of accessing product labeling using the company’s web page for the drug, a company toll-free number, and a magazine ad using five-point scale questions: “Would you say you are very unlikely, somewhat unlikely, neither unlikely nor likely, somewhat likely, or very likely to use: (insert method)?” We also asked this question in the context of if privacy was a concern to determine if their preferences changed.

Ability to Obtain Product Labeling Using Adequate Provision Delivery Methods.

We measured ability to access the product labeling using the company’s web page, a company toll-free number, and a magazine ad by asking, “Would you say you are not able, somewhat able, or completely able to obtain the full product information for a prescription drug by using (insert method)?” (“not able”/“somewhat able”/“completely able”).

Willingness to Obtain Product Labeling Using Adequate Provision Delivery Methods.

We assessed willingness to access the product labeling using the company’s web page, a company toll-free number, and a magazine ad by asking participants, “Would you say you are unwilling, somewhat willing, or completely willing to obtain the full product information for a prescription drug by using (insert method)?” (“unwilling”/“somewhat willing”/“completely willing”).

Preference for Accessing Product Labeling in the Case that Print Magazine Ads were Unavailable.

We measured preferences for accessing the product labeling in the specific case that print ads were unavailable. The question read: “If print ads were not available and you wanted to obtain additional product information about a prescription drug, would you use (insert method)?” (“no”/“yes”). In addition to the adequate provision methods suggested by FDA guidance, participants were asked if they would use “some other source” and asked to provide open-ended answers in response.

Privacy Concerns.

To assess privacy concerns we asked participants, “When seeking information about a prescription drug you are considering taking, how concerned are you about your privacy?” Privacy concerns were measured using a five-point scale from 1 = “not at all concerned” to 5 = “very concerned.” We also asked participants, “If privacy were a concern for you when getting full prescription drug product information, would you say you are very unlikely, somewhat unlikely, neither unlikely nor likely, somewhat likely, or very likely to use (insert method)?” We also asked participants, “In the case that privacy were a concern for you and you called the toll-free phone number to get the full prescription drug product information, which way would you most prefer to receive the information?” (“a customer service agent to read the information to you”/“the information to be communicated through a pre-recorded message”/“request the information be sent by mail”).

Weighting

Table 1 contains the unweighted distribution of the final 1,486 participants.

Table 1.

Weighted participant demographics

Category
	Estimated Frequency	Estimated Percentage
Age groups
18 to 59	322	21.6%
60 to 69	309	20.8%
70 to 79	398	26.8%
80 and Over	458	30.8%
Sex
Female	855	57.6%
Male	630	42.4%
Race *
American Indian or Alaska Native	38	2.6%
Asian**	5	0.4%
Black or African American	264	18.2%
Native Hawaiian and Other Pacific Islander**	1	0.1%
White	1,100	75.9%
Other	42	2.9%
Hispanic
No	1,308	90.3%
Yes	140	9.7%
Education
Less than high school	358	24.3%
High school degree	661	45.0%
Some college, no degree	240	16.3%
Associate degree or higher	211	14.4%
Income
Less than $20,000	646	49.0%
$20,000 to $40,000	380	28.8%
$40,000 to $60,000	178	13.5%
More than $60,000	114	8.6%
Browsed Internet in last month
No	1,445	97.2%
Yes	41	2.8%
If Internet browsed in last month, how many times?
Two or fewer	41	100%
Three or more	0	0%

Note that all frequencies were rounded to the nearest whole person. As such, percentages may not add to 100% and frequencies may not add to N = 1,486 due to rounding and weighting.

^*Note that race categories are not mutually exclusive.

^**Note that several race categories had fewer than 10 participants. Categories with less than 10 participants were recoded into the Other category and excluded from analyses.

Data were weighted to reflect the national population of adults age 18 or older who do not use the Internet. The sample was balanced to match estimates for the population without Internet access and without a smartphone derived from the 2017 American Community Survey (ACS; Ruggles et al., 2017). We trimmed weights at the 2nd and 98th percentiles to keep the effect low and to prevent individual respondents from having too much influence on survey estimates. An overall weight was generated for the full sample based on population parameters.

Due to an error in programming logic, we removed 513 total cases from the data set. Given that these cases may have been systematically different from other cases, a weight adjustment was implemented. Among the 513 replacement respondents, 172 cases that were equivalent in terms of skip pattern logic to the 513 removals were weighted up to represent the same proportion as the total 685 cases (original 513 plus new 172). Weights were then standardized to have a mean of 1. The standardization maintains the proportionality of the weights although the constant weight factor may change.

We used the overall weight for any statistical tests that did not involve the two questions related to the re-fielding. For analyses involving each impacted question, a specialized weight was used. The results reported below are estimates based on the weighting procedures.

Results

Internet, Phone, and Print Magazine Ad Accessibility

Less than half of the participants had the Internet available at home (n = 538, 36.2%) or outside the home (n = 585, 39.4%). Nearly all participants had either a landline or cell phone available (n = 1,441, 97.1%) and a slight majority of participants had access to print magazine ads (n = 748, 50.4%). These findings suggest that overall, low- and non-Internet users have more access to phone lines than magazine ads or the Internet.

Internet Accessibility.

Internet accessibility differed by education and income level. A larger proportion of participants with a high school education or less reported not having the Internet available at home compared to participants with higher education levels, X² (3, N = 1,465) = 17.87, p = .04. More participants with an income of less than $40,000 reported not having the Internet available at home compared to participants with income more than $40,000, X² (3, N = 1,314) = 41.44, p < .001.

Landline or Cell Phone Accessibility.

A higher proportion of older participants did not have a telephone available compared to relatively younger participants, X² (3, N = 1,485) = 26.72, p = .004.

Print Magazine Ad Accessibility.

Fewer participants who identified as either “American Indian or Alaska Native” or “Black or African American” reported that magazine ads were available compared to White participants, X² (2, N = 1,438) = 26.24, p = .001. In addition, a smaller proportion of participants who identified as Hispanic reported that magazine ads were available to them compared to non-Hispanic participants, X² (1, N = 1,438) = 8.16, p = .04. Fewer participants with less than high school education reported that magazine ads were available to them compared to participants with an education level of high school degree or higher, X² (3, N = 1,458) = 35.37, p = .001.

Ability to Use Internet, Phone, and Print Magazine Ads

Few participants reported having complete ability to use the Internet at home either without help (n = 155, 10.5%) or with help from another person (n = 134, 10.3%). Similarly, few participants reported having complete ability to use the Internet outside of their homes without help (n = 159, 10.8%) or with help from another person (n = 112, 8.5%). A majority of participants reported having complete ability to use a print magazine ad (n = 884, 59.5%). Most participants reported being completely able to use either a landline or cell phone (n = 1,270, 85.5%).

Ability to Use the Internet.

Ability to use the Internet differed by education, age, and income. A larger proportion of participants with a high school education or less reported not having the ability to use the Internet at home without help from another person compared to participants with higher education levels, X² (6, N = 1,446) = 54.34, p < .001. A larger proportion of participants who were at least 60 years old reported not having the ability to use the Internet at home without help from another person compared to relatively younger participants, X² (6, N = 1462) = 22.70, p < .05. In addition, a larger proportion of participants with an income of less than $20,000 reported not having the ability to use the Internet at home without help from another person compared to participants with higher incomes, X² (6, N = 1,299) = 70.61, p < .001. Similarly, a larger proportion of participants with an income of less than $20,000 reported not having the ability to use the Internet at home with help from another person compared to participants with higher incomes.

Ability to Use a Landline or Cell Phone.

Chi-square results showed a significant relationship between ability to use a telephone and race, ethnicity, education, and income level. A larger proportion of participants who identified as White reported having complete ability to use a telephone compared to American Indian or Alaska Native and Black or African American participants, X² (6, N = 1,486) = 74.21, p < .001. A smaller proportion of participants who identified as American Indian or Alaska Native reported no ability to use a telephone compared to Black or African American and White participants. There was also a significant relationship between ethnicity and ability to use a telephone, X² (2, N = 1,448) = 64.50, p < .001. A larger proportion of non-Hispanic participants reported having complete ability to use a telephone compared to Hispanic participants. Chi-square tests also showed that a larger proportion of participants with a high school degree or less reported no ability to use a telephone compared to higher educational levels, X² (5, N = 1,469) = 62.90, p < .001. Finally, a larger proportion of participants with an income less than $20,000 reported having the ability to use a telephone compared to participants with higher incomes, X² (6, N = 1,317) = 20.82, p < .001.

Ability to Use Print Magazine Ads.

Ability to use print magazine ads differed by race, ethnicity, education, age, and income. A larger proportion of participants who identified as White reported having complete ability to use magazines ads compared to American Indian or Alaska Native and Black or African American participants, X² (4, N = 1,443) = 37.44, p < .001. A smaller proportion of participants who identified as American Indian or Alaska Native reported no ability to use magazines ads compared to Black or African American participants and White participants. There was also a significant relationship pertaining to ethnicity, X² (2, N = 1,443) = 54.04, p < .001. A larger proportion of non-Hispanic participants reported no ability to use magazines ads compared to Hispanic participants.

Chi-square tests showed that a larger proportion of participants with less than high school education reported no ability to use magazines ads compared to participants with higher levels of education. There was a significant relationship between ability to use magazine ads and education, X² (6, N = 1,464) = 106.87, p = .001.

There was also a significant relationship between ability to use magazine ads and age, X² (6, N = 1480) = 35.13, p = .02. A smaller proportion of participants less than 60 years old reported having no ability to use magazine ads compared to older participants (>60 years old). There was also a significant relationship between ability to use magazine ads and income, X² (5, N = 1,340) = 39.26, p = .005. A larger proportion of participants with an income less than $20,000 reported having no ability to use magazine ads compared to participants with higher incomes (> $20,000).

Likelihood of Accessing Product Labeling Using Adequate Provision Delivery Methods

We assessed likelihood of using each adequate provision delivery method by asking participants how likely they would be to use each method if they wanted to get information about a prescription drug after seeing an advertisement on television. Overall, participants reported being somewhat unlikely to use any adequate provision method investigated in this study. Participants reporting being unlikely to use a company’s web page either with (M = 1.76, SE = .11) or without help (M = 2.24, SE = .11). Participants were also unlikely to use a magazine ad (M = 2.06, SE = .05). Of the adequate provision methods, respondents were most likely to use a toll-free number (M = 2.38, SE = .06) to obtain more information about a prescription drug.

Ability to Obtain Product Labeling Using Adequate Provision Delivery Methods

Next, we assessed ability to obtain the product labeling specifically using each adequate provision method by asking participants if they would be able to find full product information for a prescription drug using each adequate provision source.

Less than one-third of participants said they are completely able to use a company’s web page to find drug information without help (n = 194, 30.7%) and fewer reported having complete ability to use the web page with help from another person (n = 157, 26.7%). More than one-third of participants (n = 566, 38.6%) said they are completely able to use a magazine ad to obtain prescription drug information. More participants reported being completely able to find full product information for a prescription drug using a toll-free number (n = 647, 44.3%).

Willingness to Obtain Product Labeling Using Adequate Provision Delivery Methods

Few participants reported being completely willing to use the Internet to find drug information either without help (n = 168, 11.4%) or with help from another person (n = 152, 11.6%). More participants (n =256, 17.3%), comparatively, were completely willing to use a magazine ad to access prescription drug information. Approximately one-quarter of participants reported being completely willing to use a toll-free phone number to obtain full product information about a prescription drug (n = 402, 27.3%).

We used a regression model to predict willingness of participants to search for prescription drug information using each adequate provision source. Age, ability to use the Internet, technical ability, and privacy concerns were significant predictors of willingness to use a company’s web page without help from another person (see Table 2a). Post-hoc analyses revealed younger participants were more willing to use a company’s web page without help from another person to obtain prescription drug information, b = 0.41, t(458) = 3.93, p < .001. There was also a positive relationship between having the ability to use the Internet outside the home without help and willingness to use a company’s web page to obtain information about a prescription drug, b = 0.17, t(458) = 2.97, p = .003. Those participants with a higher technical ability to use the company’s web page to obtain prescription drug information, specifically, were more willing to do so, b = 0.38, t(458) = 7.68, p < .001. Participants who were more concerned about privacy were more willing to use a company’s web page without help from another person to obtain prescription drug information, b = 0.05, t(458) = 2.36, p = .02.

Table 2a.

Influence of demographics, availability, ability, technical ability, and privacy concerns on willingness to use a company’s web page without help from another person to obtain information about a prescription drug (R² = .36)

Source	df1	df2	Wald F	Sig.
(Corrected Model)	16	443	14.87	<.001
(Intercept)	1	458	4.25	.040
Age	3	456	7.50	.000
Race	2	457	2.15	.117
Ethnicity	1	458	1.61	.206
Income	3	456	.336	.799
Side effects	1	458	.563	.453
Availability of Internet at home	1	458	.376	.540
Availability of Internet outside home	1	458	.034	.855
Ability to use Internet at home without help	1	458	.046	.830
Ability to use the Internet outside home without help	1	458	8.80	.003
Ability to use company’s web page to obtain drug information without help	1	458	58.95	<.001
Privacy concerns	1	458	5.56	.019

Age, ability to use the Internet, technical ability, and privacy concerns were significant predictors of willingness to use a company’s web page with help from another person (see Table 2b). Post-hoc analyses revealed that participants who are younger were more willing to use a company’s webpage with help from another person to obtain prescription drug information, b = 0.37, t(257) = 2.40, p = .02. There was a positive relationship between ability to use the Internet outside the home with help and willingness to use a company’s web page to obtain information about a prescription drug, b = 0.19, t(259) = 2.01, p < .05. Similarly, participants with a higher technical ability to use the Internet to obtain prescription drug information were more willing to do so, b = 0.45, t(259) = 6.20, p < .001. Concern for privacy was positively associated with willingness to use a company’s web page with help from another person to obtain prescription drug information, b = 0.09, t(259) = 2.78, p = .006.

Table 2b.

Influence of demographics, availability, ability, technical ability, and privacy concerns on willingness to use a company’s web page with help from another person to obtain information about a prescription drug (R² = .38)

Source	df1	df2	Wald F	Sig.
(Corrected Model)	16	244	7.80	<.001
(Intercept)	1	259	1.49	.223
Age	3	257	3.19	.024
Race	2	258	.504	.604
Ethnicity	1	259	.173	.678
Income	3	257	.332	.802
Side effects	1	259	2.63	.106
Availability of Internet at home	1	259	2.02	.156
Availability of Internet outside home	1	259	2.82	.095
Ability to use Internet at home with help	1	259	.616	.433
Ability to use Internet outside home with help	1	259	4.03	.046
Ability to use company’s web page to obtain drug information with help	1	259	38.44	<.001
Privacy concerns	1	259	7.75	.006

Age, income, availability, technical ability, and privacy concerns were significant predictors of willingness to use a company’s toll-free number to obtain prescription drug information (see Table 2c). Post hoc analyses revealed participants who are younger were more willing to use a company’s toll-free phone number to obtain prescription drug information, b = 0.33, t(1,183) = 3.60, p < .001. Participants with less than $40,000 yearly income were more likely to use a toll-free phone number, b = 0.22, t(1,183) = 2.13, p = .03. Technical ability to use a toll-free phone number positively predicted willingness to use a toll-free phone number to obtain prescription drug information, b = 0.44, t(1,183) = 13.59, p < .001. Participants who were more concerned about privacy were more willing to use a company’s toll-free phone to obtain prescription drug information, b = 0.07, t(1,183) = 3.87, p < .001.

Table 2c.

Influence of demographics, availability, ability, technical ability, and privacy concerns on willingness to use a company’s toll-free phone number to obtain information about a prescription drug (R² = .27)

Source	df1	df2	Wald F	Sig.
(Corrected Model)	16	1168	25.28	<.001
(Intercept)	1	1183	2.53	.112
Age	3	1181	6.92	.000
Race	2	1182	.783	.457
Ethnicity	1	1183	.896	.344
Income	3	1181	2.73	.043
Side effects	1	1183	1.75	.187
Availability of landline phone	1	1183	.269	.604
Availability of cell phone	1	1183	1.48	.224
Ability to use landline phone	1	1183	.142	.707
Ability to use cell phone	1	1183	.273	.601
Ability to use company’s toll-free phone number to obtain drug information	1	1183	184.54	<.001
Privacy concerns	1	1183	14.995	<.001

Age, availability, technical ability, and privacy concerns were significant predictors of willingness to use a company’s print ad in a magazine to obtain prescription drug information (see Table 2d). Post hoc analyses revealed participants who are younger were more willing to use a magazine ad to obtain prescription drug information, b = 0.23, t(1,190) = 2.51, p = .01. Availability of magazines positively predicted willingness to use magazine ads to find prescription drug information, b = 0.24, t(1,190) = 3.77, p < .001. Participants with more technical ability to use a magazine ad were more willing to use magazines to obtain prescription drug information, b = 0.32, t(1,190) = 8.27, p < .001. Participants who were more concerned about privacy were more willing to use a magazine ad to obtain prescription drug information, b = 0.05, t(1,190) = 2.95, p = .003.

Table 2d.

Influence of demographics, availability, ability, technical ability, and privacy concerns on willingness to use a company’s print ad in a magazine to obtain information about a prescription drug (R² = .25)

Source	df1	df2	Wald F	Sig.
(Corrected Model)	14	1177	18.28	<.001
(Intercept)	1	1190	8.69	.003
Age	3	1188	3.82	.010
Race	2	1189	.469	.626
Ethnicity	1	1190	.640	.424
Income	3	1188	.547	.650
Side effects	1	1190	.248	.619
Availability of print ad in a magazine	1	1190	14.20	.000
Ability to use print ad in a magazine	1	1190	3.25	.072
Ability to use company’s print ad in a magazine	1	1190	68.51	<.001
Privacy concerns	1	1190	8.69	.003

Preference for Accessing Product Labeling if Print Magazine Ads were Unavailable

We asked participants about preferences for using the various adequate provision sources if print magazine ads were unavailable. A majority of participants reported that they would not use the company’s web page without help (n = 1,216, 82.6%) or with help from another person (n = 916, 74.5%). More participants indicated that they would use a toll-free number (n = 612, 41.7%) or some other source (n = 744, 50.9%) to find information if print ads are not available.

Open-ended answers for participants who said they would use “some other source” were coded for if they mentioned each alternative source. A majority of participants who chose “some other source” mentioned a doctor or nurse (n = 444, 59.1%). An estimated one-quarter of participants said that they would consult a pharmacist (n = 201, 26.7%). Smaller proportions of participants reported preferring to look at drug packaging (n = 41, 5.5%), consulting a friend or family member without medical expertise (n = 34, 4.5%), or consulting a friend of family member with medical expertise (n = 8, 1.1%).

Privacy Concerns

Participants reported some concern about privacy when seeking prescription drug information (M = 3.34, SE = 0.6). Participants were asked the likelihood of using various adequate provision methods if privacy were a concern. A paired samples t-test compared the likelihood of using each adequate provision method when privacy was a concern to participants’ self-reported likelihood of using each adequate provision without considering privacy. Overall, participants were somewhat unlikely to use any of the adequate provision methods if privacy were a concern. Participants reported that they would be least likely to use the company’s web page with help from another person (M = 1.87, SE = .06) and without help from another person (M = 1.93, SE = .06) when privacy was a concern. Neither of these results were significantly different than when participants were not considering privacy.

Participants reported that they would be most likely to use a toll-free phone number to obtain drug information if privacy were a concern (M = 2.67, SE = .06). The likelihood of using a toll-free number increased from reported likelihood when participants did not consider privacy (M = 2.40, SE = 1.63), t(1,448) = 5.99, p < .001. Similarly, participants were more likely to use magazine ads when privacy was concerned (M = 2.20, SE = .06) compared to when participants were not asked to think about privacy concerns (M = 2.11, SE = 1.49), t(1,446) = 2.17, p < .03. Participants were also asked to indicate which adequate provision method they would most prefer to use if privacy were a concern and they called a toll-free number for more information. A majority of participants (n = 883, 64.4%) reported that they would prefer to request information be sent by mail when privacy is a concern. The second most preferred option was a customer service agent reading the information (n = 396, 28.9%) whereas the least preferred option was listening to the information through a pre-recorded message (n = 92, 6.7%).

Discussion

This research assessed utilization of adequate provision in prescription drug broadcast ads among a nationally representative sample of low- and non-Internet users, including a Spanish-language subsample. This is an important issue because in recent years, an assumption has begun to emerge that consumers expect and prefer to access product labeling information online. However, with an estimated 10% of people in the United States being non-Internet users (Pew Research Center, 2019a), alternatives to the Internet likely remain critical. The present research sheds light on questions of access, ability, likelihood, willingness, and preference for utilizing the various sources of adequate provision among the offline population.

Our findings reveal nearly universal access to phone lines (approximately 97% of participants have access). However, less than half of respondents reported that they were “completely able” to use a toll-free number to obtain prescription drug information. When asked about print magazine ads, half of our participants reported them as accessible, and slightly less than half reported being “completely able” to use a magazine ad specifically. Not surprisingly, the Internet was reported as least accessible to this population, with about one-third of participants reporting accessibility, and about the same proportion reporting being able to access a drug company’s webpage specifically. Regarding willingness to access these various sources of adequate provision, the largest proportion of participants reported being “completely willing” to access the toll-free number.

These findings show that the offline population tends to view the toll-free number as most accessible and most aligned with what they have the capability and willingness to utilize (RQ1). Nearly all low- or non-Internet users have access to a phone, whereas only about half have access to print magazine ads and one-third have access to the Internet. Although only a minority of participants report being “completely able” and “completely willing” to use a toll-free number to get product information, these reported rates are much higher than what was observed for the other sources.

Although print magazine ads are accessible to many offline individuals, and many individuals report having the capability to use them, this source of adequate provision appears redundant with what is offered by the toll-free number, which is both much more accessible and similar in reported capability to use as well as higher in reported willingness to use (RQ2). We also note that existing FDA guidance advises providing individuals who call the toll-free number with the option to receive printed materials via mail (FDA, 1999). Therefore, concomitant publication of a print ad may be redundant with the toll-free number for offline individuals when such individuals have greater access to the toll-free number and additionally may receive printed, mailed materials by request through the toll-free number.

Using the Internet to obtain product labeling online if they have a specific information-seeking goal and if the information is only available online would pose a challenging task for low- and non-Internet users. We observed low reported access to the Internet and very few of these individuals report any ability to use the Internet, a finding which held true both inside and outside the home, and with or without help (RQ3). Finally, participants reported low willingness to access a company’s webpage for a drug, including in the specific case that a print ad was unavailable. In all, participants reported low access, ability, and willingness to use Internet sources to access the full product labeling, again supporting the notion that provision of non-Internet sources of adequate provision is critical for this population.

Participants in our sample expressed some concern about privacy when seeking this information (RQ4). More practically, in cases where privacy was a concern, participants reported low likelihood of seeking out the product labeling across the various sources of adequate provision. Participants in our sample reported being more likely to use the toll-free number than other sources of adequate provision in cases where privacy was a concern. In this specific case of calling the toll-free number, participants reported a preference for having printed materials mailed to them. Thus, the preference among privacy-conscious low- and non-Internet users is to call the toll-free number and request that printed materials be mailed to them. This finding is a bit surprising given that requesting mailed materials requires that one provide their name and address, which inherently compromises privacy. Nonetheless, these findings lend further support to the notion that for offline individuals, print ads may no longer offer unique value relative to the toll-free number.

Findings from the present research may serve to inform future policy development; for example, evaluating certain elements of FDA’s existing guidance pertaining to adequate provision (FDA, 1999). Most apparent is the expectation for the distribution of print materials independent of the toll-free number. Whereas consumers in the past may have relied heavily on print ads, brochures, and other similar materials (Aikin, Swasy, & Braman, 2004), today’s consumers generally favor use of the Internet (Aikin et al. in press) to access health information, assuming they are able to access and use the Internet; and otherwise they tend to favor the toll-free number (as found in the present research) which can provide the same information either directly or via mailed print materials by request. Existing guidance also cites privacy concerns as a rationale for recommending publication of print materials independent of the toll-free number. However, participants in our research reported greater willingness to utilize the toll-free number in cases where privacy was a concern. In response to these findings, we recommend that guidance recommendations be revised to avoid suggestion that print materials independent of the toll-free number offer unique value relative to other methods of adequate provision.

At the same time, certain elements of FDA’s existing guidance should be maintained. There remains a sizeable segment of the population that does not use the Internet, and inaccessibility to the Internet, or inability to use the Internet, are known contributors to health disparities such as those observed among certain racial and ethnic minorities (Brodie et al., 2000). Thus, to ensure fair access to full product information among Internet-disadvantaged populations, it is critical that offline methods for accessing product labeling remain available. FDA’s guidance already advises that sources of adequate provision extend to those without the Internet (FDA, 1999).

In conclusion, the present research reveals some important considerations regarding adequate provision utilization among low- and non-Internet users. Most apparent is that reliance on Internet sources exclusively is inadequate. Among the remaining sources of adequate provision, the toll-free number appears to offer the most practical value to the offline population compared to other discussed approaches, and although print advertising can also offer value, it does not appear to offer unique value relative to what is offered by the toll-free number, particularly when the toll-free number incorporates an option to have printed materials mailed to the consumer.