Printer Friendly

Risks, benefits, and generic substitution.

Increased substitution of generic for brand-name drug products has long been advocated as a means of decreasing consumers' expenditures on prescription medicines. Available evidence indicates that individual consumers realize major savings when they use generic products (Bloom, Wierz, and Pauly 1986). However, aggregate consumer savings fall short of potential savings because the incidence of substitution is low (Chain Drug Review 1988; 1989).

All 50 states have passed legislation which permits pharmacists to substitute generic products on prescriptions written for brand-name drugs. While consumers and prescribers have the legal right to request or deny substitution, for the great majority of prescriptions they leave the choice to the pharmacist. Thus, for the most part, pharmacists determine the extent to which generic substitution will occur.

Pharmacists, and the pharmacies at which they practice, may receive several types of benefits as a result of substituting generic for branded products. Many benefits accrue from price differences between generic and brand-name products. Generic drugs typically cost pharmacies 50 to 80 percent less than brand-name counterparts. Large cost differences allow pharmacies to sell generic products at prices which are substantially less than those charged for brand-name products, and which yield larger profit margins (in both dollar and percentage terms) to the pharmacy (Bloom, Wierz, and Pauly 1986). By using generic products pharmacies can improve both profitability and price image. This, in turn, may affect the amount of time pharmacists must spend dealing with consumer complaints about prescription prices, contribute to increased consumer loyalty to the pharmacy, and allow pharmacists to feel good about saving consumers money on prescriptions. Pharmacists may also feel that selecting the product to dispense, rather than dispensing the exact product specified by the prescriber, enhances their professional status.

Pharmacists may also incur certain risks when they dispense generic substitutes. Carroll, Siridhara, and Fincham (1986) examined the relationship between perceived risks and the extent to which pharmacists dispensed generic products. They used a commonly cited model of perceived risk (Bearden and Mason 1978; Cunningham 1967; Peter and Tarpey 1975; Peter and Ryan 1976) which included several different dimensions of risk, or types of losses, which might affect a pharmacist's decision to dispense a generic drug product. Their model included dimensions of financial, social, psychological, performance, and convenience risks. Results indicated that perceived risk was inversely related to the rate at which pharmacists substituted generic for brand-name drugs. Responding pharmacists were most concerned with performance and psychological risks.

The objective of this study was to expand upon work begun in an earlier study. The first of these was to correct two methodological shortcomings. The earlier study used a measure of perceived risk which was calculated as the sum of the products of the probability of each dimension of risk multiplied by the importance of that dimension. Mathematically, the multiplicative model can be expressed as follows:

[Mathematical Expression Omitted]

where PR is overall perceived risk, PL is probability of loss, IL is importance of loss, and n is the dimensions of risk. While the multiplicative model has been frequently used (Bearden and Mason 1978; Cunningham 1967; Peter and Tarpey 1975; Peter and Ryan 1976), several researchers have noted potential problems with it (Bettman 1975; Peter and Ryan 1976). Schmidt (1973) noted that multiplication of attitude scales similar to the ones used in measuring perceived risk was "not a theoretically meaningful operation." He found that transformations of such scales which involved only moving the location of the zero point resulted in large changes in correlation coefficients. Because of these potential problems, a nonmultiplicative model was employed.

The second shortcoming concerned generalizability of results. In the earlier study, only pharmacists from Georgia were surveyed. This study employed a larger, more representative sample. In addition, effects of perceived benefits on pharmacists' substitution behavior were examined. It seemed reasonable to assume that pharmacists' decision should be affected by expected benefits as well as by expected risks.

Specific objectives of this study were to examine the relationship between perceived risks and benefits and pharmacists' substitution behavior and to determine the most influential risks and benefits.


Data were collected via a mail survey. A mailing list of 1,007 pharmacists was purchased from a commercial vendor. The mailing list was compiled by random selection from lists maintained by the Boards of Pharmacy in Kentucky, New York, Mississippi, California, and Texas. The survey procedure consisted of an initial mailing to all pharmacists, a postcard reminder mailed one week later, and a second copy of the questionnaire mailed to nonrespondents three weeks later.

Responses were received from 501 pharmacists. Of the total questionnaires mailed, 162 were returned because of incorrect addresses. This yielded a response rate of 59 percent of delivered questionnaires. Of the responses received, 401 were complete and usable and 100 were from respondents no longer engaged in the practice of pharmacy.

Sample and population demographics were compared to assess whether the sample was representative of the population. Population data were taken from the National Center for Health Statistics publication, Characteristics of Pharmacists, United States (1984). This document is based on a census of pharmacists conducted during 1978-1979. While these data are old, they are the most current available. Chi-square tests were used to compare sample and population distributions of pharmacists' positions and places of employment. No significant differences were found (Table 1). About 20 percent of both sample and population consisted of inactive pharmacists. The sample inlcuded 26 percent women while the population included only 16 percent. This probably reflects the increasing proportions of females graduating from pharmacy schools during the last decade rather than lack of representativeness of the sample. Overall, comparisons indicated that the sample was demographically similar to the population.

The questionnaire contained a list of 24 multiple source, frequently prescribed brand-name drug products. Pharmacists were instructed to indicate whether they would dispense the brand indicated or make a substitution when presented with prescriptions for each product. They were asked to assume that substitution was permitted by prescribers and that consumers were paying for prescriptions (i.e., no third-party payer, such as Medicaid or private insurance, was involved). A


substitution rate was calculated for each respondent as the number of products for which a substitute would be dispensed.

The 24 products were selected from Pharmacy Times' "Top 200" list of most frequently prescribed drugs (Pharmacy Times 1987). Products were chosen to insure that a broad range of therapeutic categories were included. Products selected are listed in Table 2.

The questionnaire included items to measure probability and importance of five dimensions of risk: financial, social, psychological, performance, and time. Perceived benefit items were developed to measure probability and importance of time, financial, psychological, and social dimensions. No items were included to measure performance benefits because of the minimal likelihood that pharmacists would experience such benefits from use of generic products. Each item cited a dimension-specific way in which generic substitution might increase risk or benefit. For example, the item that measured probability of financial risk asked pharmacists to indicate the likelihood that increased substitution would lead to financial loss due to a law suit.

Probabilities and importances were measured with five-point ratings scales anchored by "Very Likely" and "Very Unlikely" or "Very Important" and "Very Unimportant." The midpoint of each scale was labeled "Undecided." Tables 3, 4, 5, and 6 list specific items included to measure risk and benefit perceptions.


The questionnaire was tested on a small number of pharmacists enrolled in the graduate program at the University of Georgia and on a sample of 100 pharmacists practicing in the state of Georgia. This resulted in minor changes to the questionnaire.

Multivariate analysis of variance (MANOVA) was used to determine the relationship between substitution rates and perceived risks and benefits. Respondents were divided into three groups based on their rate of generic substitution. The substitution rates selected to divide the sample were chosen to yield three groups of approximately equal size. The high substitution group consisted of pharmacists who indicated they would dispense generic substitutes for 21 or more products, the low substitution group consisted of those who indicated they would dispense substitutes for 16 or less products, and the medium substitution group consisted of the remaining respondents.

Four MANOVAs were run. The independent variable in each was substitution group. Dependent variables were, for the first MANOVA, dimensions of probability of risk; for the second MANOVA, dimensions of importance of risk; for the third, dimensions of probability of benefit; and for the fourth, dimensions of importance of benefit.


The typical respondent dispensed a substitute for 17.5 (SD = 5.6) of the 24 products. The number of substitutions per respondent ranged from 0 to 24 and the mode was 22.

Mean scores for perceived probabilities of risk ranged from 2.4 for financial risk to 3.4 for time and psychological risks (Table 3). MANOVA indicated that perceived probability of risk was significantly different across the three substitution groups (Wilks' criterion; F = 4.03, df = 10, 776, p = .001). ANOVAs indicated that each individual dimension of risk was significantly different across groups. Two canonical correlates were statistically significant. Total structure coefficients (which can be interpreted in the same manner as factor loadings) indicated moderate to high correlations between all five dimensions of risk and the first canonical correlate. The highest correlations involved performance and psychological risks. Total structure coefficients showed moderate correlations between the second canonical correlate and dimensions of time, psychological, and social risk. An examination of mean differences among substitution groups indicated that the largest absolute differences occurred for performance and psychological risks.

Mean scores for perceived importance of risk ranged from 3.4 for social risk to 4.2 for performance and psychological risk (Table 4). The MANOVA indicated no significant differences across groups in perceived importance (Wilks' criterion; F = 1.02, df = 10, 760, p = .428).

Mean scores for perceived probability of benefits ranged from 3.0 for social benefit to 4.1 for financial benefit (Table 5). Both financial and psychological benefits were perceived as at least somewhat likely


to occur. The MANOVA indicated significant differences across the three substitution groups (Wilks' criterion; F = 2.54, df = 10, 774, p = .005). ANOVAs indicated that perceived probability of financial, psychological, and social benefits differed significantly across groups.

The first canonical correlate was statistically significant. Total canonical structure coefficients indicated moderate to high correlations between the first canonical correlate and all dimensions of perceived benefit. Perceived social benefit had a particularly high structure coefficient. This indicated that perceived probability of enhanced social status was an important discriminator among groups. However, mean scores for the substitution groups on this item only ranged from 2.7 to 3.2. This indicated that none of the groups viewed enhanced social status as a likely result of generic substitution.

Mean scores for perceived importance of benefits ranged from 3.8 for time benefits to 4.2 for psychological benefits (Table 6). MANOVA incicated no significant differences across groups (Wilks' criterion; F = 1.72, df = 10, 774, p = .070).




Pharmacists' perceptionms of risks and benefits of generic substitution were significantly related to their substitution decisions. Perceptions of importance of risks and benefits were not significantly different among the three substitution groups. Perceptions of likelihood that risks and benefits would occur did differ. Consequently, pharmacists' substitution behavior appears to be related to their perceptions of the probabilities that losses and gains will occur, but not to the perceived importance of the losses and gains.

Pharmacists perceived the most risk from dimensions related to product quality--from use of products which were not biologically equivalent to products prescribed and from worries related to quality of generic products. They indicated that these losses were at least somewhat important and that there was some (although not great) likelihood that they would result from generic substitution. Those who felt losses were more likely to occur were less likely to substitute. Pharmacists felt that financial losses resulting from legal liability related to substitution would be important, but unlikely to occur. They perceived little risk of social losses.

Pharmacists perceived the greatest benefits of generic substitution from dimensions related to cost savings. They saw likely and important benefits from increased profits and from feeling good about saving consumers money. Pharmacists who felt benefits were more likely to occur were more likely to substitute.

The results suggest that efforts to increase the incidence of substitution should focus on two areas: quality and cost. Pharmacists' fears of increased use of non-bioequivalent products, and their worries about it, could be minimized by assuring pharmacists of the quality and equivalence of generic products. This might be accomplished, for example, by promotional messages which feature comparative bioavailability data, cite manufacturers' lack of product recall problems, or stress manufacturers' quality control efforts. Advertisements by several generic manufacturers have stressed these points. The brand-name industry has also been aware of the importance of the quality issue. Several companies have mounted anti-generic promotional campaigns aimed at creating doubts about the equivalence of generic products. Such campaigns clearly focus on performance and psychological risks. Results of this study underscore the importance of the Food and Drug Administration's (FDA) efforts to stop such advertising.

State and federal regulatory efforts should be most successful if aimed at assuring and certifying quality and equivalence of generic products. This could be accomplished through making published lists of bioequivalent products--such as the FDA's "Orange Book"--more accessible to practicing pharmacists. Mandating the FDA approval or some type of FDA certification of bioequivalence be documented in product packaging should also help.

Efforts to increase substitution should also focus on cost savings potential of generic products. There is little more generic manufacturers can do in this area. Generic drugs' cost savings are already well known. Since enactment of the 1984 Waxman-Hatch Act, price competition within the generic industry has intensified. This has resulted in even lower prices for generic products.

Many states have laws which require pharmacists to pass substitution-related cost savings on to consumers. (In this context, cost savings refer to the difference between the price the pharmacy pays for a brand-name product and the price it pays for an equivalent generic product.) Such requirements may hinder substitution. Because increased profits are seen as a major benefit of substitution, any action which limits profitability may inhibit substitution. Further, because pharmacists see saving consumers money as a major benefit, they are likely to pass some portion of cost savings to consumers even when not required by state law. Hence, it is unlikely that cost savings pass-on legislation will actually increase total consumerr savings on prescription drugs.

Some third-party payers--such as private insurance plans and state Medicaid agencies--have recently attempted to boost substitution rates by offering financial incentives to pharmacies. These efforts usually involve paying pharmacies a higher dispensing fee for a generic product than for its branded counterpart. The results suggest that financial incentives should increase use of generic products.

In summary, results suggest that pharmacists are most concerned with performance-related risks and cost-related benefits of generic substitution. High and low substituting pharmacists differ little in the importance which they attribute to these factors. They differ substantially in their views of the likelihood that gains and losses will occur. Consequently, the results indicate that efforts to increase pharmacists' use of generic products should focus on increasing the likelihood that substitution will result in greater profits and consumer savings and on decreasing the likelihood that non-bioequivalent products will be used.


Bearden, William O. and J. Barry Mason (1978), "Consumer-Perceived Risk and Attitudes Toward Generically Prescribed Drugs," Journal of Applied Psychology, 63: 741-746.

Bettman, James R. (1975), "Information Integration in Consumer Risk Perception: A Comparison of Two Models of Component Conceptualization," Journal of Applied Psychology, 60: 381-385.

Bloom, Bernard S., David J. Wierz, and Mark V. Pauly (1986), Cost and Price of Comparable Branded and Generic Pharmaceuticals," Journal of the American Medical Association, 256(November 14): 2523-2530.

Carroll, Norman V., Chanaporn Siridhara, and Jack E. Fincham (1986), "Perceived Risks and Pharmacists' Generic Substitution Behavior," The Journal of Consumer Affairs, 20(Summer): 36-47.

Chain Drug Review (1988), "Several factors hike generic drug sales," 10(January 18): 28-29.

Chain Drug Review (1989), "CNS drugs fueling penetration," 11(January 16): 12.

Cunningham, Scott M. (1967), "The Major Dimensions of Perceived Risk," in Risk Taking and Information Handling in Consumer Behavior, D. F. Cox (ed.), Boston, MA: Graduate School of Business Administration, Harvard University.

National Center for Health Statistics (1984), Characteristics of Pharmacists, United States, Washington, DC: U.S. Government Printing Office.

Peter, J. Paul and Lawrence X. Tarpey, Sr. (1975), "A Comparative Analysis of Three Consumer Decision Strategies," Journal of Consumer Research, 2(June): 29-37.

Peter, J. Paul and Michael J. Ryan (1976), "An Investigation of Perceived Risk at the Brand Level," Journal of Marketing Research, 13(May): 184-188.

Pharmacy Times (1987), "Top 200 Drugs of 1986: A New Record: Antibiotics Account for Seven of the Top New Generic Prescriptions," 53(April): 32-39.

Schmidt, Frank L. (1973), "Implications of a Measurement Problem for Expectancy Theory Research," Organizational Behavior and Human Performance, 10: 243-251.

Norman V. Carroll is Associate Professor of Pharmacy Administration, School of Pharmacy, Virginia Commonwealth University, Richmond, VA; and Alan P. Wolfgang is Assistant Professor, Department of Pharmacy Care Administration, University of Georgia, Athens, GA.

This work was supported by the Eckerds Drug Company.
COPYRIGHT 1991 American Council on Consumer Interests
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:prescription drugs
Author:Carroll, Norman V.; Wolfgang, Alan P.
Publication:Journal of Consumer Affairs
Date:Jun 22, 1991
Previous Article:Misperceptions of grocery advertising.
Next Article:Spousal ethical justifications of casino gambling: a psychometric analysis.

Related Articles
Generic vs. brand name.
Express Scripts Study Shows Substantial Savings Opportunities for Consumers, States, Health Care Purchasers With Generics, Says GPhA.
Keep track of your drugs.
The diagnosis on generics.
Epilepsy drugs worries; Mailbag.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters