- Email: [email protected]
Pharmacy students' opinions of direct-to-consumer advertising: A pilot study at one university
Research in Social and Administrative Pharmacy 9 (2013) 458–466
Original Research
Pharmacy students’ opinions of direct-to-consumer advertising: A pilot study at one university Amanda R. Harrington, B.S.a,*, Shane P. Desselle, R.Ph., Ph.D.b, David A. Apgar, Pharm.D.a, Elizabeth Hesselbacher, Pharm.D.a, Aaron Pie´, Pharm.D.a, Aimee Quesnel, Pharm.D.a, Terri L. Warholak, R.Ph., Ph.D.a a
Department of Pharmacy Practice and Science, University of Arizona, 1295 North Martin Avenue, Tucson, AZ 85721, USA b College of Pharmacy, California Northstate University, 10811 International Drive, Rancho Cordova, CA 95670, USA
Abstract Background: Direct-to-consumer advertisement (DTCA) of prescription medications has become an important informational source for health care consumers. As future health care professionals on the front line of potential communication and dispensing of products emerging from DTCA, it is important to elicit the attitudes of student-pharmacists. Objectives: This study aims to (1) evaluate the validity of the DTCA attitudinal questionnaire using Rasch rating scale analysis and (2) investigate the attitudes of pharmacy students toward DTCA and determine whether these attitudes were associated with years of pharmacy education and demographic characteristics. Methods: This investigation used a cross-sectional print-based questionnaire to evaluate the attitudes of pharmacy students toward DTCA of prescription medications. The 16-item questionnaire included items addressing the attitudes of pharmacy students toward DTCA with respect to patients’ knowledge of medications, pharmacists’ interaction with patients, and overall consumer judgment of medical prescriptions. Analyses included Rasch analysis and a multiple linear regression. Results: A total of 243 students submitted usable questionnaires (85% response rate). Item response categories were collapsed from 5 categories to 3, and 4 items were removed to achieve acceptable Rasch model fit. Pharmacy students demonstrated little difficulty in agreeing with the statements suggesting that DTCA helps patients take a more active role in health care and had the most difficulty in agreeing with items suggesting that DTCA may lead to inappropriate prescribing to satisfy patient requests. Students’ overall support for DTCA was the only variable that predicted the questionnaire score (P ! .001). Conclusions: In conclusion, the Rasch analysis evaluated the psychometric properties of the instrument and identified the necessity to adapt the questionnaire from previous iterations to adequately fit the student population. Future research should examine factors that contribute to the variance in attitudes toward DTCA among a larger and more heterogeneous population. Ó 2013 Elsevier Inc. All rights reserved. Keywords: Direct-to-consumer advertising; Pharmacy student; Rasch; Attitude
* Corresponding author. Tel.: þ1 480 495 8519; fax: þ1 520 626 2466. E-mail address: [email protected] (A.R. Harrington). 1551-7411/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.sapharm.2012.07.005
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Introduction Direct-to-consumer advertising (DTCA) as a form of pharmaceutical marketing was first allowed in the United States in 1985.1 Since the origin of DTCA in the U.S. advertising market, the advertising regulations have become progressively less stringent and led to a significant discourse regarding the impact of DTCA on the health care system.2 The United States and New Zealand are currently the only developed countries to allow DTCA for pharmaceutical marketing.3 Although DTCA is currently banned in Canada and the European Union, DTCA is a topic of debate and legal battles, fueled by pressure from pharmaceutical companies to allow DTCA.4 In support of DTCA, it is argued that the advertisement increases public awareness and education. Several studies have focused on 2 key areas that challenge this argument.5-9 First, advertisements often provide inadequate information regarding treatment risks, alternatives, and preventative measures. Second, patients are unlikely to understand the information presented in advertisements because health literacy levels are often lower than the level needed to evaluate the medical information advertised. A lack of adequate information presented about pharmaceutical products coupled with a potential lack of understanding render the value of DTCA questionable.10 Also, the source of this information is the companies that sell the products. The companies’ private sector interest is quite different from that of a public sector sponsored public service address. It has been suggested that the benefit of DTCA is not the educational value but rather the motivational aspect that encourages patients to seek care for conditions that they may have previously ignored because they were unaware of available treatment options. Studies have demonstrated that DTCA has an impact on prescribing patterns and the perceived patient-physician relationship.10-13 However, DTCA-motivated medication requests have contributed to inappropriate prescribing and an increased overall use of medication.13 In the continuum of patient care, pharmacists play a crucial role in delivering, optimizing, and monitoring pharmacotherapy, in addition to serving as a link between the patient and the physician. Pharmacists also serve as an educational resource for patients. Both the American Pharmacists Association and the Pharmaceutical Research and Manufacturers of America recognize
459
the potential of DTCA to increase the awareness of disease and educate people on treatment options. Both organizations stress the importance that advertised pharmaceutical information must be accurate, not misleading, and be consistent with Food and Drug Administration (FDA)approved labeling.14,15 Although these leading pharmacy organizations have released DTCA statements, sparse information is available regarding pharmacists’ perspective on pharmaceutical advertising practices. Available research on pharmacists’ opinions toward DTCA is inconclusive.16 One study found that pharmacists believe that DTCA can be beneficial, whereas other studies have found that they believe DTCA to have detrimental effects.17,18 Pharmacy students are among the next generation of heath care providers and, therefore, it is of interest to investigate their views toward DTCA. Naik et al.19 elicited the opinions of pharmacy students toward DTCA and found that although students had an overall negative attitude toward televised and printed advertising that did not use a professional labeling format, they had a positive attitude toward print advertisement presented with a professional labeling format (ie, using technical medical terminology to provide detailed information on treatment side effects and contraindications). Although informative, Naik et al. examined attitudes toward specific advertisements, in particular, for drugs with therapeutic alternatives going over-the-counter (ie, protein pump inhibitors and antihistamine). This contrasts with the research-examining practitioner opinions toward the DTCA phenomenon in general.17,18 Additionally, an update on professional pharmacy students’ views is apropos because with the passage of time, contemporary students would hardly recall a time wherein DTCA advertisements were not pervasive. Finally, the instrument used by Naik et al., as well as many of those used to gather opinions of practitioners, was subjected to reliability analysis and face validity but has not been evaluated strictly for construct and criterion-related validation. The objectives of this study were to: (1) evaluate the validity of the DTCA attitudinal questionnaire using a Rasch rating scale analysis and (2) investigate the attitudes of students from a College of Pharmacy toward DTCA and determine whether these attitudes were associated with years of pharmacy education and certain demographic characteristics.
460
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Methods Study design and participants This investigation used a cross-sectional printbased questionnaire to evaluate the attitudes of pharmacy students toward DTCA of prescription medications. Students were eligible to participate if they were enrolled in their first, second, or third year of didactic study in the College of Pharmacy. Two hundred seventy-five students were given the opportunity to participate in the study. The attitudinal assessment questionnaire was distributed during a lecture period of a required pharmacy course, during which participating students voluntarily elected to complete the instrument. Students absent from the lecture period were not included in the study. The University Institutional Review Board approved this study. Attitudinal questionnaire The attitudinal questionnaire included 16 items selected from a 24-item instrument created by Desselle and Aparasu17 to determine pharmacists’ support of DTCA of prescription medicines.17 The 16 items were selected after consulting the investigators regarding the items that were thought to best fit the study’s target population (ie, pharmacy students). In addition to these 16 items, 2 more were included; the first elicited pharmacy students’ overall opinion toward DTCA, whereas the second item addressed whether students who had participated as pharmacy interns had been exposed to DTCA during their internship experience. The questionnaire included items addressing the attitudes of pharmacy students toward DTCA with respect to patients’ knowledge of medications, pharmacists’ interaction with patients, and overall consumer judgment of medical prescriptions. Study participants rated their level of agreement with the 16 questionnaire items on a 5-point Likert rating scale with response options that included “strongly disagree,” “disagree,” “neutral,” “agree,” and “strongly agree.” The final 2 categorical questionnaire items assessed whether the student personally supported DTCA and if they had ever received a request for DTCA information from a patient. For these items, the students were given the option to respond “yes,” “no,” or an alternate response if the yes or no answer did not apply (ie, “do not know” or “not applicable”). The same Likert scale was used for both negatively and positively worded questionnaire items. To ensure consistent scoring
during item summation, reverse scoring was used to convert all negative item scores to positive. That is, if a participant selected “strongly disagree” for a negative item, it was reversed to “strongly agree” during item scoring. Rasch evaluation of instrument validity To achieve construct inference from the data, one must demonstrate that the measurement model produces linear measures, overcomes missing data, gives estimates of precision, includes devices to detect misfitting data, and separates the parameters of the object being measured from the measurement instrument.20 A Rasch measurement model is able to assess these issues and evaluate questionnaire validity, including assessing construct variance and construct representation. The Rasch measurement process assesses potential threats to validity to support the expected relationships (ie, level of DTCA agreement) with characteristics implied by the construct of the questionnaire.20-24 Once the results of a Rasch analysis have determined the questionnaire items that fit the model, the items qualify for subsequent summation (ie, that the items are true interval-level data). The single score for a questionnaire item produced from the summation of participant responses may be used as an indicator of meaningfulness of an underlying questionnaire attribute. Furthermore, summative scoring indicates that the objective evidence is available to demonstrate that all items measure the same construct, produce additivity of measures, and the probability of a participant responding to 1 item is not dependent on another instrument item.25,26 A full description of Rasch methodology and application was beyond the scope of this study, and additional more exhaustive articles are available for further reading.20-22,25-27 A Rasch analysis was performed to determine if the modified 16-item questionnaire was able to measure the attitudinal concept it was intended to measure when used in the student sample. The Rasch rating scale model was selected for the data analysis because it offers objective evidence of unidimensionality (ie, that all items measure the same construct) and yields additivity of measures necessary for the subsequent analytic procedures. Specifically, it allows for the transformation of ordinal-level data into interval-level data on a log odds or logit measure of scale if the data fit the model.20 Data analysis Standard procedures were used to evaluate construct variance and representation using
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Rasch analysis.27 Data were initially entered into an Excel file and then imported into Winsteps, version 3.71.0.1,28 to conduct a Rasch rating scale analysis. To assess the unidimensionality of the questionnaire, the infit and outfit statistics yielded from the Rasch analysis were used to examine whether the data fit the Rasch model. The infit statistic is the weighted mean square residual evaluating the inlying range of expected values that represent the difference between the observed and expected values of items that have a difficulty level similar to the ability of the subject. The outfit statistics is the unweighted mean square residual reflecting the outlying range expected values that represent the items that are far from the subject’s ability level. The farther the distance between the expected value and the measured fit statistic, the more the item misfits the Rasch model.29 The Rasch analysis was used to evaluate the content and construct validity criteria of subsequent conclusions deduced from the results. Content validity refers to the degree to which the instrument items represent the content it is designed to measure, whereas construct validity represents the degree to which the questionnaire tool measures a theoretical construct (ie, DTCA).30 Items were removed if they did not meet the Rasch item criteria and were, therefore, not used in the subsequent analyses.20 Z tests of Rasch item scores were performed to assess item gap significance.29 Once the scale items had been assessed to determine if a transformation to interval-level data was appropriate, a multiple linear regression was conducted to assess the influence of selected independent variables on the Rasch score. The dependent variable was the score summation of those items retained in the questionnaire after the Rasch analysis, representing the attitude of students-pharmacists toward DTCA. Instrument items comprising the dependent variable included demographics, experiential background, and future career interests. Stata version was used to calculate descriptive statistics and perform all post-Rasch analyses.31 An alpha level of 0.05 for regression covariates was chosen a priori.
Results Demographics Of the 275 pharmacy students eligible to complete the survey, 234 students submitted usable questionnaires (85% response rate). Questionnaires were included in the analysis if the
461
student answered at least 80% of the items. More than 50% of the respondents for each didactic year were females. Most of the students in their first or second year of pharmacy school had not received a degree before starting the professional program, whereas most of the students in their third didactic year had a prior degree. Most students had some experience in the hospital setting before completing the questionnaire and expressed interest in working in an independent pharmacy setting after the completion of their pharmacy degree. All demographic information is listed in Table 1. DTCA questionnaire Construct variance For a summative item score to be used as an indicator of how pharmacy students view DTCA, it must first be demonstrated that the items work together psychometrically. Psychometric properties are required for an instrument to meet the standards of a Rasch analysis. To evaluate whether the questionnaire fits the Rasch model, Linacre’s rating scale category guidelines were used to assess the rating scale-based data for category frequency, ordering, rating-to-measure inferential coherence, and the quality of the scale from measurement and statistical procedures.32 Guidelines used to assess the rating scale were the following: (1) at least 10 observations per category, (2) regular observation distribution, (3) average measures advance monotonically with category, (4) outfit mean squares less than 2, (5) step calibration advance, (6) ratings imply measures and measure imply ratings, and (7) step difficulties advance 1.4 logits or more but less than 5 logits. Rasch analysis was used to determine if collapsing categories yielded a better fit model. The categories “strongly disagree,” “disagree,” “neutral,” “agree,” and “strongly disagree” were collapsed to “disagree,” “neutral,” and “agree.” After comparing the models based on Linacre guidelines, the model with collapsed categories was a better fit model and was the final model used for the results. Rasch analysis also was performed to assess the extent to which each questionnaire item contributed to the model fit. The following 4 items were found to exceed the acceptable Rasch rating scale model fit criteria and were excluded from the final analysis: item 1, “DTCA increases awareness of medications among patients”; item 6, “DTCA stimulates patients to see their health
462
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Table 1 Demographic information Information
Didactic year in pharmacy school Year 1 (N ¼ 92), n (%)
Male 53 (42.4) Age, y !30 84 (91.3) 30-39 7 (7.6) R40 1 (1.1) Highest degree received Associate 10 (10.9) Bachelors 5 (5.4) Graduate 30 (32.6) None 47 (51.1) Pharmacy practice setting with the most experience Chain 8 (8.7) Hospital 53 (57.6) Independent 27 (29.3) Other 2 (20) None 2 (2.2) Desired pharmacy practice setting after graduation Chain 30 (32.6) Hospital 13 (14.1) Independent 33 (35.9) Other 7 (7.6) Undecided 9 (9.8)
care provider more often”; item 10, “DTCA increases patient knowledge of available treatment alternatives”; and item 16, “Disagreements between pharmacists and consumers may arise because of DTCA.” Remaining items in the questionnaire comprising the dependent variable for the regression analysis are listed in Fig. 1. Construct representation The observed average item scores for the student respondents are depicted in Fig. 1. This figure demonstrates how this sample used the categories disagree, neutral, and agree to respond to the questionnaire items. Items on the right-hand side of the figure are ordered based on how difficult it was for the students to agree with that item. The items are listed in ascending order, in which students had little difficulty agreeing with question 3, whereas question 8 was the most difficult. Multiple linear regression model Covariates in the regression model predicting the outcome score of the questionnaire included age, sex, and student support for DTCA. Model coefficients are listed in Table 2. The adjusted R2 value was 6.5%. Student support for DTCA was
Year 2 (N ¼ 76), n (%)
Year 3 (N ¼ 66), n (%)
29 (38.7)
22 (33.3)
68 (91.9) 6 (8.1) 0 (0)
46 (69.7) 14 (21.2) 6 (9.1)
6 1 29 39
(7.9) (1.3) (38.2) (51.3)
7 1 34 24
(10.6) (1.5) (51.5) (36.4)
5 51 11 6 2
(6.7) (68) (14.7) (60) (2.7)
2 50 9 2 3
(3) (75.8) (13.6) (20) (4.5)
24 16 27 7 2
(34.8) (21.1) (35.5) (9.2) (2.6)
15 12 32 4 3
(22.7) (18.2) (48.5) (6.1) (4.5)
the only statistically significant variable to predict the questionnaire score (P ! .001). The statistical contribution of the global question on support for DTCA provides evidence for criterion-related validity of the items comprising the attitudinal questionnaire.
Discussion Opinions about DTCA have pervaded the scientific literature for well over a decade. The impact of DTCA, although, cannot be neatly categorized as “good” or “bad,”33 and the overall net effect is difficult to quantify.34,35 The US FDA may eventually mandate long-advocated changes in advertisement regulations to ensure optimal drug benefit and improve risk information retention among consumers, particularly those with compromised health literacy.36 As such, it is important that health care practitioners remain abreast of developments in DTCA, understand that it is a phenomenon with which they must deal, and use any opportunity to educate patients in a professional manner.37 The current study examined the likelihood that future pharmacists agree with certain commonly
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
463
Fig. 1. Observed average item scores. Reverse scoring was used to convert negative scores to positive for items 2, 5, 8, 9, 12, and 14 (eg, if a participant selected “strongly disagree” for a negative item, it was reversed to strongly agree during item scoring).
believed effects of DTCA, which provides some indication of how they might in the future answer patient inquiries about medications advertised. The student-pharmacists surveyed agreed more easily with items suggesting that DTCA helps patients to take a more active role in health care and seek more information from pharmacists. Students also more easily agreed with the item pertaining to DTCA improving the role of pharmacists as health educators, despite their concomitant agreement with pharmacists’ advice being relied on less as a result of DTCA. Indeed, it would bode well for patients if pharmacists were to embrace patient self-management and seize any
opportunity to use their expertise, all the while demonstrating and understanding that patients will seek and acquire information about drugs from a variety of sources. At the same time, student-pharmacist respondents demonstrated greater difficulty agreeing with statements that DTCA may lead to inappropriate prescribing, promotes patients’ overreliance on medications, and results in increased drug costs. This corresponds with the evidence to suggest that physicians often heed to the requests of patients desiring a specific drug but not when the drug is inappropriate.38 Additionally, any attempt to pinpoint the effects of DTCA on
Table 2 Multiple linear regression model Covariates
Beta coefficient
Standard error
Statistical significance
Age Sex Student support for DTCA Desired pharmacy practice setting after graduation
0.014 0.093 2.085 0.069
0.91 0.801 0.475 0.302
0.99 0.91 0.000a 0.819
a
Statistically significant at alpha ¼ 0.05 level.
464
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
consumer cost and reliance on medications is difficult and depends on many factors, including the therapeutic class, patient access, availability of therapeutic alternatives, and effectiveness of advertising campaigns.39,40 Overall, the attitudes of student-pharmacists toward DTCA in the current study were relatively neutral to positive. This is in contrast to the results of Naik et al.19 and to another study assessing the attitudes of pharmacists.41 There could be several reasons for these differences. First, the current study precluded items that did not perform well psychometrically in a Rasch analysis, and as such, the measuring instrument was at least unique. Second, the passing of time wherein students and practitioners have grown more accustomed to DTCA and to growing patient autonomy may have attenuated negative predispositions, as might have the lack of overt evidence of truly deleterious impacts of DTCA. Recent years have seen a decline in the use of celebrities in advertisements concurrently with more favorable news coverage of DTCA, both of which may fuel more positive affective responses from students and health professionals.42,43 It also is possible that students in one university setting might have a different view than those in another, should any courses at either institution stress the more positive or the more negative aspects of DTCA. After years of DTCA, faculty might be presenting DTCA in a more nuanced manner, allowing for discussions of social cognitive theory and other complex phenomena to explain consumers’ behavior resulting from the exposure to medication information.44,45 Although much can be gleaned from the examination of student-pharmacists’ opinions of various aspects of DTCA, we still do not know much about what governs these opinions. First, 4 items from the original 16-item questionnaire were removed because they did not adequately meet the requirements of the Rasch analysis for the transformation to interval-level data. The Rasch analysis procedure is a quantitative method used to assess an ordinal-level item and determine whether it may be used as interval level. It is of benefit to transform the data if it is of interest to use the variables in subsequent analyses (eg, multiple linear regression). Items may have been excluded because of a lack of clarity or misinterpretation. It is also possible that the items removed are a part of another dimension, and further research is required to assess other constructs within which these items fit. Moreover,
pharmacy students did not have a baseline reference of patient knowledge and, therefore, were not equipped to evaluate a change in patient knowledge. Second, the regression analysis explained only 6.5% of the variance in responses, and this was accounted for mostly by a global item measure of overall support for the concept. Student-pharmacist age, sex, year in the program, or prior experience with DTCA explained an inconsequential amount of the variance. It is important to understand the source of variation of opinions toward DTCA because this helps to govern any educational interventions and identify particular factors governing the way that practitioners and student-pharmacists handle the opportunities and challenges presented to them as a result of DTCA. Two sources that might account for more variance in attitudes toward DTCA are additional attitudinal questionnaire items and other explanatory variables. The Rasch analysis pointed to the need to eliminate certain items from the originally proposed instrument, which contributed in part to some of the gaps found within the measuring instrument. Attitudes toward DTCA, as a construct, might comprise factors in addition to the ones measured, including beliefs about medications in general, perceptions in the veracity of myriad sources of health information, DTCA’s effect on patient relationships with other health care providers, business/profit opportunities and challenges resulting from DTCA, and any perceived pressure from keeping up with the DTCA advertisement campaigns. Future research might involve inclusion of additional items for testing. Additionally, the effect of other independent variables could be assessed. These might include attitudes toward pharmaceutical manufacturers, any of various political stances, previous experiences from patients with requests for information about an advertised product, the results of any personal request for a medication advertised, attitudes about marketing/advertising in general, and even one’s own health locus of control. Social networking and ethical stances on the interplay of developing technologies and consumerism may also influence the perception of DTCA.46 Future research can investigate what effects these might have. There were several strengths and limitations inherent in the current study. Strengths include use of the Rasch model, the student population, and instrument selection. Ensuring that the instrument adequately met the Rasch model requirements provided greater confidence internally
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
in the results yielded from the multiple linear regression analysis. The study population included students from multiple years in the University’s pharmacy curriculum, minimizing biases inherent from a particular course or instructor. The instrument selected for this study was previously validated on pharmacists with the intention of comparing results over time. Limitations to the study include self-reporting, less than perfect psychometric performance of the instrument, and a limited population sample. The fit of the instrument to the Rasch model was adequate, but its psychometric performance was not exemplary, thus leaving room for improvement. Last, the student population was gathered from a single College of Pharmacy. Further studies can assess attitudes toward DTCA in a larger more diverse population of student-pharmacists and practitioners to increase the generalizability of the questionnaire results.
Conclusions This study reports the opinions of studentpharmacists toward various aspects of prescription drug DTCA. As future health care professionals on the front line of potential communication and dispensing of products and services emanating from DTCA, it is important to elicit the attitudes of student-pharmacists about this phenomenon, in general, rather than toward a specific advertisement. The student-pharmacist respondents demonstrated little difficulty in agreeing with statements suggesting that DTCA helps patients to take a more active role in health care and presents pharmacists with opportunities as health educators. As demonstrated with a multiple linear regression model, pharmacy education and demographic characteristics evaluated were found to have no significant predictive effect on the attitudes of pharmacy students toward DTCA. A Rasch analysis evidenced the psychometric properties of the instrument and the necessity to adapt it from previous iterations by removing certain poorly performing items. Future research should examine the factors that contribute to the variance in attitudes toward DTCA among a larger and more heterogeneous population. Acknowledgments No funding was provided for this study.
465
References 1. Lyles A. Direct marketing of pharmaceuticals to consumers. Annu Rev Public Health 2002;23(1): 73–91. 2. Code of Federal Regulations: Food and Drugs; 2001. National Archives and Records Administration. CFR title 21, Vol. 4. Available at: http://www. access.gpo.gov/nara/cfr/waisidx_01/21cfr202_01.html. Accessed 30.01.12. 3. Hoek J, Maubach N. Consumers’ knowledge, perceptions, and responsiveness to direct-to-consumer advertising of prescription medicines. N Z Med J 2007;120(1249):U2425. 4. Guthrie P. Direct-to-consumer-advertising debated in the United States and European Union. CMAJ 2007;176(10):1404. 5. Bell RA, Wilkes MS, Kravitz RL. The educational value of consumer-targeted prescription drug print advertising. J Fam Pract 2000;49(12):1092–1098. 6. Chao BA. Evaluating the educational content of direct-to-consumer fulfillment materials. Am J Health Syst Pharm 2005;62(6):620–625. 7. Curry TJ, Jarosch J, Pacholok S. Are direct to consumer advertisements of prescription drugs educational?: comparing 1992 to 2002. J Drug Educ 2005; 35(3):217–232. 8. Frosch DL, Krueger PM, Hornik RC, Cronholm PF, Barg FK. Creating demand for prescription drugs: a content analysis of television direct-to-consumer advertising. Ann Fam Med 2007;5(1):6–13. 9. Kaphingst KA, Rudd RE, DeJong W, Daltroy LH. Literacy demands of product information intended to supplement television direct-to-consumer prescription drug advertisements. Patient Educ Couns 2004;55(2):293–300. 10. Mintzes B, Barer ML, Kravitz RL, et al. How does direct-to-consumer advertising (DTCA) affect prescribing? A survey in primary care environments with and without legal DTCA. CMAJ 2003;169(5): 405–412. 11. Shah MB, Bentley JP, McCaffrey DJ, Kolassa EM. Direct-to-consumer advertising and the patientphysician relationship. Res Social Adm Pharm 2005; 1(2):211–230. 12. Zachry WM III, Dalen JE, Jackson TR. Clinicians’ responses to direct-to-consumer advertising of prescription medications. Arch Intern Med 2003;163(15): 1808–1812. 13. Kravitz RL, Epstein RM, Feldman MD, et al. Influence of patients’ requests for direct-to-consumer advertised antidepressants. JAMA 2005;293(16): 1995–2002. 14. Direct to consumer advertising: providing accurate information about disease and treatment options. Pharmaceutical Research and Manufacturers of America Web Site. Available at: http://www.phrma.org/about/ principles-guidelines/direct-consumer-advertising. Accessed 30.01.12.
466
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
15. Bough MA. Statement of the American Pharmacists Association to the Food and Drug Administration. Available at: http://www.pharmacist.com/AM/template. cfm?cx¼001477644988688360425:01rpa0bcmrq&cof¼ FORID:9&ie¼UTF-8§ion¼/AM/Template.cfm? Section¼Home2&template¼/CustomSource/google results.cfm&q¼directþtoþconsumerþadvertising&sa. x¼0&sa.y¼0&sa¼Search&siteurl¼www.pharmacist. com/. Accessed 30.01.12. 16. Maine LL. Direct-to-consumer advertising: a pharmacy perspective. Clin Ther 1998;20:C104–C110. 17. Desselle SP, Aparasu R. Attitudinal dimensions that determine pharmacists’ decisions to support DTCA of prescription medication. Drug Inf J 2000;34(1): 103–114. 18. Farthing-Papineau EC, Peak AS. Pharmacists’ perceptions of the pharmaceutical industry. Am J Health Syst Pharm 2005;62(22):2401–2409. 19. Naik RK, Borrego ME, Gupchup GV, Dodd M, Sather MR. Pharmacy students’ knowledge, attitudes, and evaluation of direct-to-consumer advertising. Am J Pharm Educ 2007;71(5):86. 20. Smith EL, Smith RF. Introduction to Rasch Measurement: Theory, Models and Applications. Maple Grove, MN: JAM Press; 2004. 21. Draugalis J, Jackson TR. Objective curricular evaluation: applying the Rasch model to a cumulative examination. Am J Pharm Educ 2004;68(2):35. 22. Jackson TR, Draugalis J, Slack MK, Zachry WM, D’Agostino J. Validation of authentic performance assessment: a process suited for Rasch modeling. Am J Pharm Educ 2002;66(3):233–242. 23. Colliver J, Williams R. Technical issues: test application. AAMC. Acad Med 1993;68(6):454–460. 24. Briggs AL, Jackson TR, Bruce S, Shapiro NL. The development and performance validation of a tool to assess patient anticoagulation knowledge. Res Soc Adm Pharm 2005;1(1):40–59. 25. Wright BD, Stone MH. Best Test Design. Kent, GB: Atlantic Books; 1979. 26. Rasch G. Probabilistic Models for Some Intelligence and Attainment Tests. Chicago, IL: University of Chicago Press; 1980. 27. Wright BD, Mok M. Rasch models overview. J Appl Meas 2000;1(1):83–106. 28. Linacre JM. WinstepsÒ Rasch Measurement Computer Program. Beaverton, OR: Winsteps.com; 2012. 29. Liao PM, Campbell SK. Examination of the item structure of the Alberta infant motor scale. Pediatr Phys Ther 2004;16(1):31–38. 30. Kerlinger FN, Lee HB. Foundations of Behavioral Research. Fort Worth, TX: Harcourt College Publishers; 2000.
31. StataCorp. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP; 2009. 32. Linacre JM. Investigating rating scale category utility. J Outcome Meas 1999;3(2):103–122. 33. Datti B, Carter MW. The effect of direct-toconsumer advertising on prescription drug use by older adults. Drugs Aging 2006;23(1):71–81. 34. Frosch D, Grande D. Direct-to-consumer advertising of prescription drugs. LDI Issue Brief 2010;15: 1–4. 35. Atherly A, Rubin PH. The cost effectiveness of direct to consumer advertising for prescription drugs. Med Care Res Rev 2009;66(6):639–657. 36. Rotelli MD, Dowsett SA, Elsner MW, Holdsworth SM, Pitts PJ. A new model for communicating risk information in direct-to-consumer print advertisements. Drug Inf J 2007;41:111–127. 37. Desselle S. Direct-to-consumer prescription drug advertising and pharmacy practice. Am J Pharm Educ 2004;68(3):1–9. 38. Weissman JS, Blumenthal D, Silk AJ, et al. Physicians report on patient encounters involving directto-consumer advertising. Suppl Web Exclusives. Health Aff (Millwood); 2004. W4-219–W4-233. 39. Norey E, Simone TM, Mousa SA. The impact of direct-to-consumer advertised drugs on drug sales in the US and New Zealand. Appl Health Econ Health Policy 2008;6(2–3):93–102. 40. Brekke KR, Kuhn M. Direct to consumer advertising in pharmaceutical markets. J Health Econ 2006; 25(1):102–130. 41. Friedman M, Gould J. Attitudes and opinions of pharmacists toward direct-to-consumer prescription drug marketing. J Pharm Finance Econ Pol 2007;16: 85–87. 42. Gerald MC. The rise and fall of celebrity promotion of prescription products in direct-to-consumer advertising. Pharm Hist 2010;52(1):13–23. 43. Hartley H, Coleman CL. News media coverage of direct-to-consumer pharmaceutical advertising: implications for countervailing powers theory. Health (London) 2008;12(1):107–132. 44. Young HN, Lipowski EE, Cline RJW. Using social cognitive theory to explain consumers’ behavioral intentions in response to direct-to-consumer prescription drug advertising. Res Social Adm Pharm 2005; 1(2):270–288. 45. Ghoshal M, Walji MF. Quality of medication information available on retail pharmacy web sites. Res Social Adm Pharm 2006;2(4):479–498. 46. McGuire AL, Diaz CM, Wang T, Hilsenbeck SG. Social networkers’ attitudes toward direct-to-consumer personal genome testing. Am J Bioeth 2009;9(6–7):3–10.
Original Research
Pharmacy students’ opinions of direct-to-consumer advertising: A pilot study at one university Amanda R. Harrington, B.S.a,*, Shane P. Desselle, R.Ph., Ph.D.b, David A. Apgar, Pharm.D.a, Elizabeth Hesselbacher, Pharm.D.a, Aaron Pie´, Pharm.D.a, Aimee Quesnel, Pharm.D.a, Terri L. Warholak, R.Ph., Ph.D.a a
Department of Pharmacy Practice and Science, University of Arizona, 1295 North Martin Avenue, Tucson, AZ 85721, USA b College of Pharmacy, California Northstate University, 10811 International Drive, Rancho Cordova, CA 95670, USA
Abstract Background: Direct-to-consumer advertisement (DTCA) of prescription medications has become an important informational source for health care consumers. As future health care professionals on the front line of potential communication and dispensing of products emerging from DTCA, it is important to elicit the attitudes of student-pharmacists. Objectives: This study aims to (1) evaluate the validity of the DTCA attitudinal questionnaire using Rasch rating scale analysis and (2) investigate the attitudes of pharmacy students toward DTCA and determine whether these attitudes were associated with years of pharmacy education and demographic characteristics. Methods: This investigation used a cross-sectional print-based questionnaire to evaluate the attitudes of pharmacy students toward DTCA of prescription medications. The 16-item questionnaire included items addressing the attitudes of pharmacy students toward DTCA with respect to patients’ knowledge of medications, pharmacists’ interaction with patients, and overall consumer judgment of medical prescriptions. Analyses included Rasch analysis and a multiple linear regression. Results: A total of 243 students submitted usable questionnaires (85% response rate). Item response categories were collapsed from 5 categories to 3, and 4 items were removed to achieve acceptable Rasch model fit. Pharmacy students demonstrated little difficulty in agreeing with the statements suggesting that DTCA helps patients take a more active role in health care and had the most difficulty in agreeing with items suggesting that DTCA may lead to inappropriate prescribing to satisfy patient requests. Students’ overall support for DTCA was the only variable that predicted the questionnaire score (P ! .001). Conclusions: In conclusion, the Rasch analysis evaluated the psychometric properties of the instrument and identified the necessity to adapt the questionnaire from previous iterations to adequately fit the student population. Future research should examine factors that contribute to the variance in attitudes toward DTCA among a larger and more heterogeneous population. Ó 2013 Elsevier Inc. All rights reserved. Keywords: Direct-to-consumer advertising; Pharmacy student; Rasch; Attitude
* Corresponding author. Tel.: þ1 480 495 8519; fax: þ1 520 626 2466. E-mail address: [email protected] (A.R. Harrington). 1551-7411/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.sapharm.2012.07.005
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Introduction Direct-to-consumer advertising (DTCA) as a form of pharmaceutical marketing was first allowed in the United States in 1985.1 Since the origin of DTCA in the U.S. advertising market, the advertising regulations have become progressively less stringent and led to a significant discourse regarding the impact of DTCA on the health care system.2 The United States and New Zealand are currently the only developed countries to allow DTCA for pharmaceutical marketing.3 Although DTCA is currently banned in Canada and the European Union, DTCA is a topic of debate and legal battles, fueled by pressure from pharmaceutical companies to allow DTCA.4 In support of DTCA, it is argued that the advertisement increases public awareness and education. Several studies have focused on 2 key areas that challenge this argument.5-9 First, advertisements often provide inadequate information regarding treatment risks, alternatives, and preventative measures. Second, patients are unlikely to understand the information presented in advertisements because health literacy levels are often lower than the level needed to evaluate the medical information advertised. A lack of adequate information presented about pharmaceutical products coupled with a potential lack of understanding render the value of DTCA questionable.10 Also, the source of this information is the companies that sell the products. The companies’ private sector interest is quite different from that of a public sector sponsored public service address. It has been suggested that the benefit of DTCA is not the educational value but rather the motivational aspect that encourages patients to seek care for conditions that they may have previously ignored because they were unaware of available treatment options. Studies have demonstrated that DTCA has an impact on prescribing patterns and the perceived patient-physician relationship.10-13 However, DTCA-motivated medication requests have contributed to inappropriate prescribing and an increased overall use of medication.13 In the continuum of patient care, pharmacists play a crucial role in delivering, optimizing, and monitoring pharmacotherapy, in addition to serving as a link between the patient and the physician. Pharmacists also serve as an educational resource for patients. Both the American Pharmacists Association and the Pharmaceutical Research and Manufacturers of America recognize
459
the potential of DTCA to increase the awareness of disease and educate people on treatment options. Both organizations stress the importance that advertised pharmaceutical information must be accurate, not misleading, and be consistent with Food and Drug Administration (FDA)approved labeling.14,15 Although these leading pharmacy organizations have released DTCA statements, sparse information is available regarding pharmacists’ perspective on pharmaceutical advertising practices. Available research on pharmacists’ opinions toward DTCA is inconclusive.16 One study found that pharmacists believe that DTCA can be beneficial, whereas other studies have found that they believe DTCA to have detrimental effects.17,18 Pharmacy students are among the next generation of heath care providers and, therefore, it is of interest to investigate their views toward DTCA. Naik et al.19 elicited the opinions of pharmacy students toward DTCA and found that although students had an overall negative attitude toward televised and printed advertising that did not use a professional labeling format, they had a positive attitude toward print advertisement presented with a professional labeling format (ie, using technical medical terminology to provide detailed information on treatment side effects and contraindications). Although informative, Naik et al. examined attitudes toward specific advertisements, in particular, for drugs with therapeutic alternatives going over-the-counter (ie, protein pump inhibitors and antihistamine). This contrasts with the research-examining practitioner opinions toward the DTCA phenomenon in general.17,18 Additionally, an update on professional pharmacy students’ views is apropos because with the passage of time, contemporary students would hardly recall a time wherein DTCA advertisements were not pervasive. Finally, the instrument used by Naik et al., as well as many of those used to gather opinions of practitioners, was subjected to reliability analysis and face validity but has not been evaluated strictly for construct and criterion-related validation. The objectives of this study were to: (1) evaluate the validity of the DTCA attitudinal questionnaire using a Rasch rating scale analysis and (2) investigate the attitudes of students from a College of Pharmacy toward DTCA and determine whether these attitudes were associated with years of pharmacy education and certain demographic characteristics.
460
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Methods Study design and participants This investigation used a cross-sectional printbased questionnaire to evaluate the attitudes of pharmacy students toward DTCA of prescription medications. Students were eligible to participate if they were enrolled in their first, second, or third year of didactic study in the College of Pharmacy. Two hundred seventy-five students were given the opportunity to participate in the study. The attitudinal assessment questionnaire was distributed during a lecture period of a required pharmacy course, during which participating students voluntarily elected to complete the instrument. Students absent from the lecture period were not included in the study. The University Institutional Review Board approved this study. Attitudinal questionnaire The attitudinal questionnaire included 16 items selected from a 24-item instrument created by Desselle and Aparasu17 to determine pharmacists’ support of DTCA of prescription medicines.17 The 16 items were selected after consulting the investigators regarding the items that were thought to best fit the study’s target population (ie, pharmacy students). In addition to these 16 items, 2 more were included; the first elicited pharmacy students’ overall opinion toward DTCA, whereas the second item addressed whether students who had participated as pharmacy interns had been exposed to DTCA during their internship experience. The questionnaire included items addressing the attitudes of pharmacy students toward DTCA with respect to patients’ knowledge of medications, pharmacists’ interaction with patients, and overall consumer judgment of medical prescriptions. Study participants rated their level of agreement with the 16 questionnaire items on a 5-point Likert rating scale with response options that included “strongly disagree,” “disagree,” “neutral,” “agree,” and “strongly agree.” The final 2 categorical questionnaire items assessed whether the student personally supported DTCA and if they had ever received a request for DTCA information from a patient. For these items, the students were given the option to respond “yes,” “no,” or an alternate response if the yes or no answer did not apply (ie, “do not know” or “not applicable”). The same Likert scale was used for both negatively and positively worded questionnaire items. To ensure consistent scoring
during item summation, reverse scoring was used to convert all negative item scores to positive. That is, if a participant selected “strongly disagree” for a negative item, it was reversed to “strongly agree” during item scoring. Rasch evaluation of instrument validity To achieve construct inference from the data, one must demonstrate that the measurement model produces linear measures, overcomes missing data, gives estimates of precision, includes devices to detect misfitting data, and separates the parameters of the object being measured from the measurement instrument.20 A Rasch measurement model is able to assess these issues and evaluate questionnaire validity, including assessing construct variance and construct representation. The Rasch measurement process assesses potential threats to validity to support the expected relationships (ie, level of DTCA agreement) with characteristics implied by the construct of the questionnaire.20-24 Once the results of a Rasch analysis have determined the questionnaire items that fit the model, the items qualify for subsequent summation (ie, that the items are true interval-level data). The single score for a questionnaire item produced from the summation of participant responses may be used as an indicator of meaningfulness of an underlying questionnaire attribute. Furthermore, summative scoring indicates that the objective evidence is available to demonstrate that all items measure the same construct, produce additivity of measures, and the probability of a participant responding to 1 item is not dependent on another instrument item.25,26 A full description of Rasch methodology and application was beyond the scope of this study, and additional more exhaustive articles are available for further reading.20-22,25-27 A Rasch analysis was performed to determine if the modified 16-item questionnaire was able to measure the attitudinal concept it was intended to measure when used in the student sample. The Rasch rating scale model was selected for the data analysis because it offers objective evidence of unidimensionality (ie, that all items measure the same construct) and yields additivity of measures necessary for the subsequent analytic procedures. Specifically, it allows for the transformation of ordinal-level data into interval-level data on a log odds or logit measure of scale if the data fit the model.20 Data analysis Standard procedures were used to evaluate construct variance and representation using
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Rasch analysis.27 Data were initially entered into an Excel file and then imported into Winsteps, version 3.71.0.1,28 to conduct a Rasch rating scale analysis. To assess the unidimensionality of the questionnaire, the infit and outfit statistics yielded from the Rasch analysis were used to examine whether the data fit the Rasch model. The infit statistic is the weighted mean square residual evaluating the inlying range of expected values that represent the difference between the observed and expected values of items that have a difficulty level similar to the ability of the subject. The outfit statistics is the unweighted mean square residual reflecting the outlying range expected values that represent the items that are far from the subject’s ability level. The farther the distance between the expected value and the measured fit statistic, the more the item misfits the Rasch model.29 The Rasch analysis was used to evaluate the content and construct validity criteria of subsequent conclusions deduced from the results. Content validity refers to the degree to which the instrument items represent the content it is designed to measure, whereas construct validity represents the degree to which the questionnaire tool measures a theoretical construct (ie, DTCA).30 Items were removed if they did not meet the Rasch item criteria and were, therefore, not used in the subsequent analyses.20 Z tests of Rasch item scores were performed to assess item gap significance.29 Once the scale items had been assessed to determine if a transformation to interval-level data was appropriate, a multiple linear regression was conducted to assess the influence of selected independent variables on the Rasch score. The dependent variable was the score summation of those items retained in the questionnaire after the Rasch analysis, representing the attitude of students-pharmacists toward DTCA. Instrument items comprising the dependent variable included demographics, experiential background, and future career interests. Stata version was used to calculate descriptive statistics and perform all post-Rasch analyses.31 An alpha level of 0.05 for regression covariates was chosen a priori.
Results Demographics Of the 275 pharmacy students eligible to complete the survey, 234 students submitted usable questionnaires (85% response rate). Questionnaires were included in the analysis if the
461
student answered at least 80% of the items. More than 50% of the respondents for each didactic year were females. Most of the students in their first or second year of pharmacy school had not received a degree before starting the professional program, whereas most of the students in their third didactic year had a prior degree. Most students had some experience in the hospital setting before completing the questionnaire and expressed interest in working in an independent pharmacy setting after the completion of their pharmacy degree. All demographic information is listed in Table 1. DTCA questionnaire Construct variance For a summative item score to be used as an indicator of how pharmacy students view DTCA, it must first be demonstrated that the items work together psychometrically. Psychometric properties are required for an instrument to meet the standards of a Rasch analysis. To evaluate whether the questionnaire fits the Rasch model, Linacre’s rating scale category guidelines were used to assess the rating scale-based data for category frequency, ordering, rating-to-measure inferential coherence, and the quality of the scale from measurement and statistical procedures.32 Guidelines used to assess the rating scale were the following: (1) at least 10 observations per category, (2) regular observation distribution, (3) average measures advance monotonically with category, (4) outfit mean squares less than 2, (5) step calibration advance, (6) ratings imply measures and measure imply ratings, and (7) step difficulties advance 1.4 logits or more but less than 5 logits. Rasch analysis was used to determine if collapsing categories yielded a better fit model. The categories “strongly disagree,” “disagree,” “neutral,” “agree,” and “strongly disagree” were collapsed to “disagree,” “neutral,” and “agree.” After comparing the models based on Linacre guidelines, the model with collapsed categories was a better fit model and was the final model used for the results. Rasch analysis also was performed to assess the extent to which each questionnaire item contributed to the model fit. The following 4 items were found to exceed the acceptable Rasch rating scale model fit criteria and were excluded from the final analysis: item 1, “DTCA increases awareness of medications among patients”; item 6, “DTCA stimulates patients to see their health
462
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
Table 1 Demographic information Information
Didactic year in pharmacy school Year 1 (N ¼ 92), n (%)
Male 53 (42.4) Age, y !30 84 (91.3) 30-39 7 (7.6) R40 1 (1.1) Highest degree received Associate 10 (10.9) Bachelors 5 (5.4) Graduate 30 (32.6) None 47 (51.1) Pharmacy practice setting with the most experience Chain 8 (8.7) Hospital 53 (57.6) Independent 27 (29.3) Other 2 (20) None 2 (2.2) Desired pharmacy practice setting after graduation Chain 30 (32.6) Hospital 13 (14.1) Independent 33 (35.9) Other 7 (7.6) Undecided 9 (9.8)
care provider more often”; item 10, “DTCA increases patient knowledge of available treatment alternatives”; and item 16, “Disagreements between pharmacists and consumers may arise because of DTCA.” Remaining items in the questionnaire comprising the dependent variable for the regression analysis are listed in Fig. 1. Construct representation The observed average item scores for the student respondents are depicted in Fig. 1. This figure demonstrates how this sample used the categories disagree, neutral, and agree to respond to the questionnaire items. Items on the right-hand side of the figure are ordered based on how difficult it was for the students to agree with that item. The items are listed in ascending order, in which students had little difficulty agreeing with question 3, whereas question 8 was the most difficult. Multiple linear regression model Covariates in the regression model predicting the outcome score of the questionnaire included age, sex, and student support for DTCA. Model coefficients are listed in Table 2. The adjusted R2 value was 6.5%. Student support for DTCA was
Year 2 (N ¼ 76), n (%)
Year 3 (N ¼ 66), n (%)
29 (38.7)
22 (33.3)
68 (91.9) 6 (8.1) 0 (0)
46 (69.7) 14 (21.2) 6 (9.1)
6 1 29 39
(7.9) (1.3) (38.2) (51.3)
7 1 34 24
(10.6) (1.5) (51.5) (36.4)
5 51 11 6 2
(6.7) (68) (14.7) (60) (2.7)
2 50 9 2 3
(3) (75.8) (13.6) (20) (4.5)
24 16 27 7 2
(34.8) (21.1) (35.5) (9.2) (2.6)
15 12 32 4 3
(22.7) (18.2) (48.5) (6.1) (4.5)
the only statistically significant variable to predict the questionnaire score (P ! .001). The statistical contribution of the global question on support for DTCA provides evidence for criterion-related validity of the items comprising the attitudinal questionnaire.
Discussion Opinions about DTCA have pervaded the scientific literature for well over a decade. The impact of DTCA, although, cannot be neatly categorized as “good” or “bad,”33 and the overall net effect is difficult to quantify.34,35 The US FDA may eventually mandate long-advocated changes in advertisement regulations to ensure optimal drug benefit and improve risk information retention among consumers, particularly those with compromised health literacy.36 As such, it is important that health care practitioners remain abreast of developments in DTCA, understand that it is a phenomenon with which they must deal, and use any opportunity to educate patients in a professional manner.37 The current study examined the likelihood that future pharmacists agree with certain commonly
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
463
Fig. 1. Observed average item scores. Reverse scoring was used to convert negative scores to positive for items 2, 5, 8, 9, 12, and 14 (eg, if a participant selected “strongly disagree” for a negative item, it was reversed to strongly agree during item scoring).
believed effects of DTCA, which provides some indication of how they might in the future answer patient inquiries about medications advertised. The student-pharmacists surveyed agreed more easily with items suggesting that DTCA helps patients to take a more active role in health care and seek more information from pharmacists. Students also more easily agreed with the item pertaining to DTCA improving the role of pharmacists as health educators, despite their concomitant agreement with pharmacists’ advice being relied on less as a result of DTCA. Indeed, it would bode well for patients if pharmacists were to embrace patient self-management and seize any
opportunity to use their expertise, all the while demonstrating and understanding that patients will seek and acquire information about drugs from a variety of sources. At the same time, student-pharmacist respondents demonstrated greater difficulty agreeing with statements that DTCA may lead to inappropriate prescribing, promotes patients’ overreliance on medications, and results in increased drug costs. This corresponds with the evidence to suggest that physicians often heed to the requests of patients desiring a specific drug but not when the drug is inappropriate.38 Additionally, any attempt to pinpoint the effects of DTCA on
Table 2 Multiple linear regression model Covariates
Beta coefficient
Standard error
Statistical significance
Age Sex Student support for DTCA Desired pharmacy practice setting after graduation
0.014 0.093 2.085 0.069
0.91 0.801 0.475 0.302
0.99 0.91 0.000a 0.819
a
Statistically significant at alpha ¼ 0.05 level.
464
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
consumer cost and reliance on medications is difficult and depends on many factors, including the therapeutic class, patient access, availability of therapeutic alternatives, and effectiveness of advertising campaigns.39,40 Overall, the attitudes of student-pharmacists toward DTCA in the current study were relatively neutral to positive. This is in contrast to the results of Naik et al.19 and to another study assessing the attitudes of pharmacists.41 There could be several reasons for these differences. First, the current study precluded items that did not perform well psychometrically in a Rasch analysis, and as such, the measuring instrument was at least unique. Second, the passing of time wherein students and practitioners have grown more accustomed to DTCA and to growing patient autonomy may have attenuated negative predispositions, as might have the lack of overt evidence of truly deleterious impacts of DTCA. Recent years have seen a decline in the use of celebrities in advertisements concurrently with more favorable news coverage of DTCA, both of which may fuel more positive affective responses from students and health professionals.42,43 It also is possible that students in one university setting might have a different view than those in another, should any courses at either institution stress the more positive or the more negative aspects of DTCA. After years of DTCA, faculty might be presenting DTCA in a more nuanced manner, allowing for discussions of social cognitive theory and other complex phenomena to explain consumers’ behavior resulting from the exposure to medication information.44,45 Although much can be gleaned from the examination of student-pharmacists’ opinions of various aspects of DTCA, we still do not know much about what governs these opinions. First, 4 items from the original 16-item questionnaire were removed because they did not adequately meet the requirements of the Rasch analysis for the transformation to interval-level data. The Rasch analysis procedure is a quantitative method used to assess an ordinal-level item and determine whether it may be used as interval level. It is of benefit to transform the data if it is of interest to use the variables in subsequent analyses (eg, multiple linear regression). Items may have been excluded because of a lack of clarity or misinterpretation. It is also possible that the items removed are a part of another dimension, and further research is required to assess other constructs within which these items fit. Moreover,
pharmacy students did not have a baseline reference of patient knowledge and, therefore, were not equipped to evaluate a change in patient knowledge. Second, the regression analysis explained only 6.5% of the variance in responses, and this was accounted for mostly by a global item measure of overall support for the concept. Student-pharmacist age, sex, year in the program, or prior experience with DTCA explained an inconsequential amount of the variance. It is important to understand the source of variation of opinions toward DTCA because this helps to govern any educational interventions and identify particular factors governing the way that practitioners and student-pharmacists handle the opportunities and challenges presented to them as a result of DTCA. Two sources that might account for more variance in attitudes toward DTCA are additional attitudinal questionnaire items and other explanatory variables. The Rasch analysis pointed to the need to eliminate certain items from the originally proposed instrument, which contributed in part to some of the gaps found within the measuring instrument. Attitudes toward DTCA, as a construct, might comprise factors in addition to the ones measured, including beliefs about medications in general, perceptions in the veracity of myriad sources of health information, DTCA’s effect on patient relationships with other health care providers, business/profit opportunities and challenges resulting from DTCA, and any perceived pressure from keeping up with the DTCA advertisement campaigns. Future research might involve inclusion of additional items for testing. Additionally, the effect of other independent variables could be assessed. These might include attitudes toward pharmaceutical manufacturers, any of various political stances, previous experiences from patients with requests for information about an advertised product, the results of any personal request for a medication advertised, attitudes about marketing/advertising in general, and even one’s own health locus of control. Social networking and ethical stances on the interplay of developing technologies and consumerism may also influence the perception of DTCA.46 Future research can investigate what effects these might have. There were several strengths and limitations inherent in the current study. Strengths include use of the Rasch model, the student population, and instrument selection. Ensuring that the instrument adequately met the Rasch model requirements provided greater confidence internally
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
in the results yielded from the multiple linear regression analysis. The study population included students from multiple years in the University’s pharmacy curriculum, minimizing biases inherent from a particular course or instructor. The instrument selected for this study was previously validated on pharmacists with the intention of comparing results over time. Limitations to the study include self-reporting, less than perfect psychometric performance of the instrument, and a limited population sample. The fit of the instrument to the Rasch model was adequate, but its psychometric performance was not exemplary, thus leaving room for improvement. Last, the student population was gathered from a single College of Pharmacy. Further studies can assess attitudes toward DTCA in a larger more diverse population of student-pharmacists and practitioners to increase the generalizability of the questionnaire results.
Conclusions This study reports the opinions of studentpharmacists toward various aspects of prescription drug DTCA. As future health care professionals on the front line of potential communication and dispensing of products and services emanating from DTCA, it is important to elicit the attitudes of student-pharmacists about this phenomenon, in general, rather than toward a specific advertisement. The student-pharmacist respondents demonstrated little difficulty in agreeing with statements suggesting that DTCA helps patients to take a more active role in health care and presents pharmacists with opportunities as health educators. As demonstrated with a multiple linear regression model, pharmacy education and demographic characteristics evaluated were found to have no significant predictive effect on the attitudes of pharmacy students toward DTCA. A Rasch analysis evidenced the psychometric properties of the instrument and the necessity to adapt it from previous iterations by removing certain poorly performing items. Future research should examine the factors that contribute to the variance in attitudes toward DTCA among a larger and more heterogeneous population. Acknowledgments No funding was provided for this study.
465
References 1. Lyles A. Direct marketing of pharmaceuticals to consumers. Annu Rev Public Health 2002;23(1): 73–91. 2. Code of Federal Regulations: Food and Drugs; 2001. National Archives and Records Administration. CFR title 21, Vol. 4. Available at: http://www. access.gpo.gov/nara/cfr/waisidx_01/21cfr202_01.html. Accessed 30.01.12. 3. Hoek J, Maubach N. Consumers’ knowledge, perceptions, and responsiveness to direct-to-consumer advertising of prescription medicines. N Z Med J 2007;120(1249):U2425. 4. Guthrie P. Direct-to-consumer-advertising debated in the United States and European Union. CMAJ 2007;176(10):1404. 5. Bell RA, Wilkes MS, Kravitz RL. The educational value of consumer-targeted prescription drug print advertising. J Fam Pract 2000;49(12):1092–1098. 6. Chao BA. Evaluating the educational content of direct-to-consumer fulfillment materials. Am J Health Syst Pharm 2005;62(6):620–625. 7. Curry TJ, Jarosch J, Pacholok S. Are direct to consumer advertisements of prescription drugs educational?: comparing 1992 to 2002. J Drug Educ 2005; 35(3):217–232. 8. Frosch DL, Krueger PM, Hornik RC, Cronholm PF, Barg FK. Creating demand for prescription drugs: a content analysis of television direct-to-consumer advertising. Ann Fam Med 2007;5(1):6–13. 9. Kaphingst KA, Rudd RE, DeJong W, Daltroy LH. Literacy demands of product information intended to supplement television direct-to-consumer prescription drug advertisements. Patient Educ Couns 2004;55(2):293–300. 10. Mintzes B, Barer ML, Kravitz RL, et al. How does direct-to-consumer advertising (DTCA) affect prescribing? A survey in primary care environments with and without legal DTCA. CMAJ 2003;169(5): 405–412. 11. Shah MB, Bentley JP, McCaffrey DJ, Kolassa EM. Direct-to-consumer advertising and the patientphysician relationship. Res Social Adm Pharm 2005; 1(2):211–230. 12. Zachry WM III, Dalen JE, Jackson TR. Clinicians’ responses to direct-to-consumer advertising of prescription medications. Arch Intern Med 2003;163(15): 1808–1812. 13. Kravitz RL, Epstein RM, Feldman MD, et al. Influence of patients’ requests for direct-to-consumer advertised antidepressants. JAMA 2005;293(16): 1995–2002. 14. Direct to consumer advertising: providing accurate information about disease and treatment options. Pharmaceutical Research and Manufacturers of America Web Site. Available at: http://www.phrma.org/about/ principles-guidelines/direct-consumer-advertising. Accessed 30.01.12.
466
Harrington et al. / Research in Social and Administrative Pharmacy 9 (2013) 458–466
15. Bough MA. Statement of the American Pharmacists Association to the Food and Drug Administration. Available at: http://www.pharmacist.com/AM/template. cfm?cx¼001477644988688360425:01rpa0bcmrq&cof¼ FORID:9&ie¼UTF-8§ion¼/AM/Template.cfm? Section¼Home2&template¼/CustomSource/google results.cfm&q¼directþtoþconsumerþadvertising&sa. x¼0&sa.y¼0&sa¼Search&siteurl¼www.pharmacist. com/. Accessed 30.01.12. 16. Maine LL. Direct-to-consumer advertising: a pharmacy perspective. Clin Ther 1998;20:C104–C110. 17. Desselle SP, Aparasu R. Attitudinal dimensions that determine pharmacists’ decisions to support DTCA of prescription medication. Drug Inf J 2000;34(1): 103–114. 18. Farthing-Papineau EC, Peak AS. Pharmacists’ perceptions of the pharmaceutical industry. Am J Health Syst Pharm 2005;62(22):2401–2409. 19. Naik RK, Borrego ME, Gupchup GV, Dodd M, Sather MR. Pharmacy students’ knowledge, attitudes, and evaluation of direct-to-consumer advertising. Am J Pharm Educ 2007;71(5):86. 20. Smith EL, Smith RF. Introduction to Rasch Measurement: Theory, Models and Applications. Maple Grove, MN: JAM Press; 2004. 21. Draugalis J, Jackson TR. Objective curricular evaluation: applying the Rasch model to a cumulative examination. Am J Pharm Educ 2004;68(2):35. 22. Jackson TR, Draugalis J, Slack MK, Zachry WM, D’Agostino J. Validation of authentic performance assessment: a process suited for Rasch modeling. Am J Pharm Educ 2002;66(3):233–242. 23. Colliver J, Williams R. Technical issues: test application. AAMC. Acad Med 1993;68(6):454–460. 24. Briggs AL, Jackson TR, Bruce S, Shapiro NL. The development and performance validation of a tool to assess patient anticoagulation knowledge. Res Soc Adm Pharm 2005;1(1):40–59. 25. Wright BD, Stone MH. Best Test Design. Kent, GB: Atlantic Books; 1979. 26. Rasch G. Probabilistic Models for Some Intelligence and Attainment Tests. Chicago, IL: University of Chicago Press; 1980. 27. Wright BD, Mok M. Rasch models overview. J Appl Meas 2000;1(1):83–106. 28. Linacre JM. WinstepsÒ Rasch Measurement Computer Program. Beaverton, OR: Winsteps.com; 2012. 29. Liao PM, Campbell SK. Examination of the item structure of the Alberta infant motor scale. Pediatr Phys Ther 2004;16(1):31–38. 30. Kerlinger FN, Lee HB. Foundations of Behavioral Research. Fort Worth, TX: Harcourt College Publishers; 2000.
31. StataCorp. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP; 2009. 32. Linacre JM. Investigating rating scale category utility. J Outcome Meas 1999;3(2):103–122. 33. Datti B, Carter MW. The effect of direct-toconsumer advertising on prescription drug use by older adults. Drugs Aging 2006;23(1):71–81. 34. Frosch D, Grande D. Direct-to-consumer advertising of prescription drugs. LDI Issue Brief 2010;15: 1–4. 35. Atherly A, Rubin PH. The cost effectiveness of direct to consumer advertising for prescription drugs. Med Care Res Rev 2009;66(6):639–657. 36. Rotelli MD, Dowsett SA, Elsner MW, Holdsworth SM, Pitts PJ. A new model for communicating risk information in direct-to-consumer print advertisements. Drug Inf J 2007;41:111–127. 37. Desselle S. Direct-to-consumer prescription drug advertising and pharmacy practice. Am J Pharm Educ 2004;68(3):1–9. 38. Weissman JS, Blumenthal D, Silk AJ, et al. Physicians report on patient encounters involving directto-consumer advertising. Suppl Web Exclusives. Health Aff (Millwood); 2004. W4-219–W4-233. 39. Norey E, Simone TM, Mousa SA. The impact of direct-to-consumer advertised drugs on drug sales in the US and New Zealand. Appl Health Econ Health Policy 2008;6(2–3):93–102. 40. Brekke KR, Kuhn M. Direct to consumer advertising in pharmaceutical markets. J Health Econ 2006; 25(1):102–130. 41. Friedman M, Gould J. Attitudes and opinions of pharmacists toward direct-to-consumer prescription drug marketing. J Pharm Finance Econ Pol 2007;16: 85–87. 42. Gerald MC. The rise and fall of celebrity promotion of prescription products in direct-to-consumer advertising. Pharm Hist 2010;52(1):13–23. 43. Hartley H, Coleman CL. News media coverage of direct-to-consumer pharmaceutical advertising: implications for countervailing powers theory. Health (London) 2008;12(1):107–132. 44. Young HN, Lipowski EE, Cline RJW. Using social cognitive theory to explain consumers’ behavioral intentions in response to direct-to-consumer prescription drug advertising. Res Social Adm Pharm 2005; 1(2):270–288. 45. Ghoshal M, Walji MF. Quality of medication information available on retail pharmacy web sites. Res Social Adm Pharm 2006;2(4):479–498. 46. McGuire AL, Diaz CM, Wang T, Hilsenbeck SG. Social networkers’ attitudes toward direct-to-consumer personal genome testing. Am J Bioeth 2009;9(6–7):3–10.