1 CTR. FOR DRUG EVALUATION & RESEARCH, FDA, GENERIC DRUGS 7, available at http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/SmallBusinessAssistance/ucm127615.pdf (last visited Feb. 7, 2011).

2 See id.; Facts About Generic Drugs, FDA, http://www.fda.gov/drugs/resourcesforyou/consumers/buyingusingmedicinesafelb/understandinggenericdrugs/ucm167991.htm (last visited Sept. 19, 2012).

3 Bapuji, Akula Thukaram et al., Bioequivalence Testing—Industry Perspective, 2 J. BIOEQUIVALENCE & BIOAVAILABILITY 98, 98 (2010)CrossRefGoogle Scholar; see also Generic Drugs; Questions and Answers, FDA, http://www.fda.gov/Drugs/ResourcesForYou/Consumers/QuestionsAnswers/ucm100100.htm (last updated Aug. 24, 2011) (stating that “[a]ccording to the Congressional Budget Office, generic drugs save consumers an estimated $8 to $10 billion a year at retail pharmacies”).

4 See AM. MED. ASSOC., SUMMARIES AND RECOMMENDATIONS OF COUNCIL ON SCIENTIFIC AFFAIRS REPORTS 13-14 (2002) (stating that in 2000, generic drugs accounted for approximately 42% of all prescriptions at the retail level in the United States); Lewis Krauskopf & Bill Berkrot, Generics to Cut U.S. Drugs Bill by $70 Billion, REUTERS, Nov. 8, 2010, available at http://www.reuters.com/assests/print?aid=ustre6a73xj-20101108 (explaining that generic prescriptions are expected to take over 85% of the prescriptions market as soon as 2014); Bioequivalence, GENERIC PHARMACEUTICAL ASS’N, http://www.gphaonline.org/issues/bioequivalence (last visited Jan. 7, 2011) (finding that in 2005, generic drugs accounted for 56% of all prescriptions dispensed in the United States. This number rose to 75% in 2010 and is estimated at 77% in 2011).

5 Krauskopf & Berkrot, supra note 4.

6 The Patient Protection and Affordable Care Act, in an effort to close the donut hole on Medicare Part D drugs, covers more brand name drugs than ever before. The efforts to close the donut hole are an exception to the overall trend of increasing generic drug use to reduce healthcare costs. See Medicare Beneficiary Savings and the Affordable Care Act, HEALTHCARE.GOV, http://www.healthcare.gov/law/resources/reports/affordablecareact.html (last visited June 23, 2012). Covered brand name drugs were discounted by manufacturers by 50% starting in 2011, with the remaining costs to be subsidized by up to 25% by 2020. Id. Generic drugs were also discounted 7% in 2011 and will rise up to 75% in 2020. Id. Indirectly, healthcare reform encourages generic use by passing the cost of healthcare onto Medicaid. Id. Individual states have laws that encourage generic drug use over brand drugs to varying degrees. See Recent Medicaid Prescription Drug Laws, 2001-2012, NAT’L CONF. OF STATE LEGS., http://www.ncsl.org/issues-research/health/medicaid-pharmaceutical-laws-and-policies.aspx (last visited June 27, 2012); Medicaid Preferred Drug Lists for Mental Health and Substance Abuse, NAT’L CONF. OF STATE LEGS., http://www.ncsl.org/documents/health/PDL-2-2012.pdf (last visited Sept. 9, 2012) (showing a chart of differing state laws regarding preferred drug lists).

7 See Welage, Lynda et al., Understanding the Scientific Issues Embedded in the Generic Drug Approval Process, 41 J. AM. PHARMACEUTICAL ASSOC. 856, 856 (2001)Google Scholar; see also Peterson, Gregory M., Generic Substitution: A Need for Clarification, 71 BRIT. J. CLINICAL PHARMACOLOGY, 966, 966 (2011)CrossRefGoogle ScholarPubMed (stating that the debate is often fueled intentionally by the pharmaceutical industry).

8 See Blix, Hege S. et al., Drugs with Narrow Therapeutic Index as Indicators in the Risk Management of Hospitalised Patients, 8 PHARMACY PRAC. 50, 50 (2010)Google ScholarPubMed; Haines, Stuart T., Substituting Warfarin Products: What's the Source of the Problem?, 45 ANNALS PHARMACOTHERAPY 807, 807 (2011).CrossRefGoogle Scholar See generally Yamada, Mikko & Welty, Timothy, Generic Substitution of Antiepileptic Drugs: A Systematic Review of Prospective and Retrospective Studies, 45 ANNALS PHARMACOTHERAPY 1406 (2011)CrossRefGoogle ScholarPubMed (finding that several retrospective studies imply a lack of BE while the prospective studies find few significant differences in pharmacokinetic measures).

9 See Blix, supra note 8, at 50.

10 See Haines, supra note 8, at 808.

11 Trinka, Eugen et al., Requirements for Generic Antiepileptic Medicines: A Clinical Perspective, 258 J. NEUROLOGY 2128, 2130-31 (2011).CrossRefGoogle ScholarPubMed

12 Pub. L. No. 75-717, 52 Stat. 1040 (1938) (codified as amended at 21 U.S.C. §§ 301-399d (2006)).

13 Katz, Russell, FDA: Evidentiary Standards for Drug Development and Approval, 1 J. AM. SOC’Y EXPERIMENTAL NEUROTHERAPEUTICS 307, 307 (2004).CrossRefGoogle ScholarPubMed

14 See Nasr, Alexander et al., Unapproved Drugs in the United States and the Food and Drug Administration, 28 ADVANCES THERAPY 842, 846-47 (2011).CrossRefGoogle ScholarPubMed

15 See id. at 847.

16 See AM. MED. ASSOC., supra note 4, at 3.

17 Id.

18 See id.

19 Welage et al., supra note 7, at 857.

20 Id.

21 Id.

22 Pub. L. No. 98-417, 98 Stat. 1585 (1984) (codified as amended at 15 U.S.C. §§ 68b-68c, 70b (2006); 21 U.S.C. §§ 301, 355, 360cc (2006); 28 U.S.C. § 2201 (2006); 35 U.S.C. §§ 156, 271, 282 (2006)).

23 AM. MED. ASSOC., supra note 4, at 3; see also Welage et al., supra note 7, at 857.

24 AM. MED. ASSOC., supra note 4, at 2.

25 21 C.F.R. § 320.1(e) (2012) (elaborating on the definition of BE); see CTR. FOR DRUG EVALUATION & RESEARCH, GUIDANCE FOR INDUSTRY: STATISTICAL APPROACHES TO ESTABLISHING BIOEQUIVALENCE (2001) [hereinafter CDER], available at http://www.fda.gov/downloads/Drugs/…/Guidances/ucm070244.pdf (an active moiety is the part of the drug molecule responsible for the physiological or pharmacological action of the drug substance).

26 See Bialer, Meir & Midha, Kamal K., Generic Products of Antiepileptics Drugs: A Perspective on Bioequivalence and Interchangeability, 51 EPILEPSIA 941, 942 (2010).CrossRefGoogle Scholar

27 See id.

28 See id.

29 Welage et al., supra note 7, at 859-62.

30 Cmax is “an indirect measure of drug absorption … and is influenced by the extent of absorption.” Id. at 861. Cmax is a weak measure of absorption rate. Id.

31 See Bialer & Midha, supra note 26, at 942.

32 Id.

33 Davit, Barbara M. et al., Comparing Generic and Innovator Drug: A Review of 12 Years of Bioequivalence Data from the United States Food and Drug Administration, 43 ANNALS PHARMACOTHERAPY 1583, 1585 (2009)CrossRefGoogle Scholar; Howland, Robert H., Evaluating the Bioavailability and Bioequivalence of Generic Medications, 48 J. PSYCHOSOCIAL NURSING & MENTAL HEALTH SERVCES 13, 14 (2010).Google ScholarPubMed

34 Howland, supra note 33, at 14.

35 Columb, Malachy O. & Lutz, Jan M., Bioequivalence and Non-Inferiority Trials, 20 CURRENT ANAESTHESIA & CRITICAL CARE 98, 99 (2009).CrossRefGoogle Scholar

36 Id.

37 Id.

38 A two-way crossover study is one where two groups receive separate drugs, and then after a period of no drugs, each group is given the other's drug. See generally Patel, Shashank et al., Single-dose, Two-way Crossover, Bioequivalence Study of Mycophenolate Mofetil 500 mg Tablet Under Fasting Conditions in Healthy Male Subjects, 33 CLINICAL THERAPEUTICS 378 (2011)CrossRefGoogle ScholarPubMed (describing a two-way crossover study to determine BE of mycophenolate). This limits variation of subjects as a cause of BE or non-BE findings. See AM. MED. ASSOC., REPORT 2 OF THE COUNCIL ON SCIENCE AND PUBLIC HEALTH (A-07) 1-2 (2007).

39 See Howland, supra note 33, at 14.

40 See Davit et al., supra note 33, at 1585.

41 See id. at 1590.

42 See id. at 1585; see also Bialer & Midha, supra note 26, at 942 (finding that variation in bioavailability of up to 10% may occur). The 90% confidence intervals for the ratio “of each pharmacokinetic variable must lie between 0.80 and 1.25.” Peterson, Gregory M., Generic Substitution: A Need for Clarification, 71 BRIT. J. CLINICAL PHARMACOLOGY 966, 966 (2011)CrossRefGoogle ScholarPubMed. This range is an indication of certainty of the study results and does not mean that the average observed Cmax and AUC ratios or plasma concentrations can vary by this range. Id. Products usually must vary by less than 10% to satisfy the 90% CI required for BE. See id.

43 See Trinka, Eugen et al., Requirements for Generic Antiepileptic Medicines: A Clinical Perspective, 258 J. NEUROLOGY 2128, 2129 (2011)CrossRefGoogle ScholarPubMed; see also Christians, Uwe et al., Bioequivalence Testing of Immunosuppressants: Concepts and Misconceptions, 77 KIDNEY INT’L 51, 51 (2010)Google Scholar.

44 Krauss, Gregory L. et al., Assessing Bioequivalence of Generic Antiepilepsy Drugs, 70 ANNALS NEUROLOGY 221, 223-24 (2011)CrossRefGoogle Scholar (stating that Cmax values varied by 15 to 25% in 11% of studies); see also AlAmeri, Mubarak et al., Generic and Therapeutic Substitutions: Are They Always Ethical in Their Own Terms?, 32 PHARMACY WORLD & SCI. 691, 693 (2010)CrossRefGoogle Scholar (citing that in Sweden between 1972 and 1996, the increase in the generic market share was associated with the increase in the number of reported side effects in seven of the fifteen medicines studied).

45 Davit et al., supra note 33, at 1584.

46 Id.

47 See id.; see also 21 C.F.R. § 320.1 (2012).

48 The FDA currently uses an average BE approach in which the term “average” refers to the comparison of population averages of a BE measure of interest. CDER, supra note 25, at 2. For immediate release drugs, a two-way crossover study is conducted on healthy adults with ages ranging from approximately twenty-four to thirty-six. See Howland, supra note 33, at 14. An example of a two-way crossover study is a research project in which group A takes the test (generic) drug and group B takes the reference (brand name) drug. See CDER, supra note 25, at 7. After an appropriate resting period between doses, the groups switch to the opposite drug from which they began. Id. This design eliminates variation between individuals because the reference and test drug are compared within the same individual. Id. For extended release drugs (drugs with longer half-lives), parallel design studies may be used to eliminate carry-over effects of the drug. Id. A parallel study is one in which group A takes one type of drug while group B takes the other type of drug. Id. The two groups do not crossover to take the opposite drug. Id. Other variations of crossover studies may also be employed. Id. However, the above-mentioned methods are most commonly used. See id.

49 See CDER, supra note 25, at 2.

50 Id.

51 See FDA Considering Tighter Controls on Generics, GENERICS & BIOSIMILARS INITIATIVE (July 1, 2011), http://www.gabionline.net/Generics/News/FDA-considering-tighter-controls-on-generics/(highlight)/FDA%20CONSIDERING%20TIGHTER%20CONTROLS%20ON%20GENERICS.

52 See Bioequivalence, supra note 4; see also Gagne, J.J. et al., Refilling and Switching of Antiepileptic Drugs and Seizure-Related Events, 88 CLINICAL PHARMACOLOGY & THERAPEUTICS 345, 349-50 (2010).CrossRefGoogle ScholarPubMed

53 See Krauss et al., supra note 44, at 224.

54 Welage et al., supra note 7, at 11.

55 See Bialer & Midha, supra note 26, at 941-42.

56 See id. at 942.

57 See id.

58 All generic, multisource drug markets should be bioequivalent and therapeutically equivalent to a single standard reference listed drug (RLD). See Bapuji, Akula Thukaram et al., Bioequivalence Testing—Industry Perspective, 2 J. BIOEQUIVALVENCE & BIOAVAILABILITY 98, 99 (2010)Google Scholar. This will help avoid significant variations among generic drugs and brand name drugs. Id. Selecting a RLD product is a constant challenge to industry. Id.

59 See Birkett, Donald, Generics—Equal or Not?, 26 AUSTL. PRESCRIBER 85, 86 (2003).CrossRefGoogle Scholar

60 See id.

61 Yamada & Welty, supra note 8, at 1407.

62 See Bialer & Midha, supra note 26, at 945.

63 Id.

64 See Arnold, Mark E., Implications of Differences in Bioanalytical Regulations Between Canada, USA and South America, 3 BIOANALYSIS, 253, 253, 257 (2011).CrossRefGoogle ScholarPubMed

65 See Birkett, supra note 59, at 86 (explaining that generic drugs can have different excipients which can sometimes alter bioavailability as was the case with phenytoin in the 1970s); see also AlAlmeri, supra note 44, at 693 (stating that “[d]ifferences in formulation have also been noted in generic versions of some drugs”).

66 Sankar, Raman et al., Understanding Therapeutic Equivalence in Epilepsy, 15 CNS SPECTRUM 112, 117 (2010).CrossRefGoogle ScholarPubMed

67 Id.

68 Refilling AEDs comes with inherent risks that are not well understood. Emergency treatment for seizures increases 2.1- to 2.3-fold even when refilling the same AED prescription from the same manufacturer. See Gagne et al., supra note 52, at 348. The estimates were the same for switching between different manufacturers except the confidence intervals were wider. Id. at 349. Patients with stable conditions who must refill undergo a series of steps, from calling in the refill, picking up the refill, and taking it without a time delay. Id. The act of refilling may cause a delay in treatment if one step of the process is delayed. See id. Also, if the pills appear different than before, patients may hesitate to continue treatment for fear of having received the wrong prescription. See id. Further tests need to be done that control for such scenarios. The act of refilling the prescription, even without a switch to generic drugs, increases the probability of a seizure. See id. The variability in bioavailability of AEDs can lead to more breakthrough seizures, even when the drug is made by the same manufacturer. See id. at 347; see also Facts about Current Good Manufacturing Practices, FDA, http://www.fda.gov/Drugs/DevelopmentApprovalProcess/Manufacturing/ucm169105.htm (last updated June 25, 2009).

69 FDA Ensures Equivalence of Generic Drugs, FDA (Aug. 2002), http://www.fda.gov/Drugs/EmergencyPreparedness/BioterrorismandDrugPreparedness/ucm134444.htm (quoting Gary Buehler, Director of FDA's Office of Generic Drugs) (“The standards for quality are the same for brand name and generic products.”).

70 See id.

71 Sankar et al., supra note 66, at 117.

72 Id.

73 See Yamada & Welty, supra note 8, at 1407.

74 Fitzgerald, Christine L. & Jacobson, Mercedes P., Generic Substitution of Levetiracetam Resulting in Increased Incidence of Breakthrough Seizures, 45 ANNALS PHARMACOTHERAPY 1, 4 (2011).CrossRefGoogle Scholar

75 Welage et al., supra note 7, at 864 (while “the designation of a product as an NTI drug does not mean that the product will have bioequivalence problems,” a majority of pharmacists reported being “uncomfortable” prescribing generic versions of certain NTI drugs, and “some clinicians have suggested that more rigid bioequivalence guidelines are needed for these types of drugs.”).

76 Sankar et al., supra note 66, at 114.

77 AlAmeri et al., supra note 44, at 693; see also Columb & Lutz, supra note 35, at 100 (noting that the absence of differences does not make two drugs identical).

78 Sankar et al., supra note 66, at 115 (explaining that certain characteristics “make bioavailability a particular concern: low water solubility, low therapeutic index, and non-linear pharmacokinetics”).

79 Many retrospective studies show a significant difference between generic and brand name drugs, while prospective studies do not support these findings. Information is scarce, and currently available studies are often conflicting and unreliable, which make the BE criteria difficult to assess. See generally Yamada & Welty, supra note 8.

80 See id. (discussing antiepileptic drugs); see also Dentali, Francesco et al., Brand Name Versus Generic Warfarin: A Systemic Review of the Literature, 31 PHARMACOTHERAPY 386 (2011)CrossRefGoogle Scholar (discussing anticoagulant drugs); Review of Therapeutic Equivalence Generic Bupropion XL 300 mg and Wellbutrin XL 300 mg, FDA (Sept. 18, 2009), http://www.fda.gov/AboutFDA/CentersOffices/OfficeofMedicalProductsandTobacco/CDER/ucm153270.htm (discussing antidepressant drugs).

81 AEDs are NTI drugs that require careful titration of each individual patient to achieve the effective dose resulting in no breakthrough seizures.. See Duh, Mei Sheng et al., The Economic Implications of Generic Substitution of Antiepileptic Drugs: A Review of Recent Evidence, 10 PHARMACOTHERAPY 2317, 2320 (2009)Google Scholar. About half of newly diagnosed epilepsy patients can gain control of their seizures through their first treatment. Id. at 2320-21. Another 20% of patients can gain control of their seizures after a change in treatment drugs. Id. at 2321. Control of occasional seizures cannot be obtained by the remaining 30% of patients seeking treatment. Id. Patients that are fortunate enough to achieve stability of their epilepsy may require polytreatment, consisting of multiple medications added to monotherapy to achieve optimal treatment. Id.

82 See id.; see also Sander, Josemir et al., Generic Substitution of Antiepilelptic Drugs, 10 NEUROTHERAPY 1887, 1887 (2010)CrossRefGoogle Scholar (stating that “[d]espite optimum treatment,” epilepsy is associated with a 20 to 30% refractory rate with current treatments).

83 Duh et al., supra note 81, at 2320.

84 There is a trend among neurologists to differentiate between BE and therapeutic equivalence. Although closely related, BE is used in place of clinical trials to ensure therapeutic equivalence. See Bialer & Midha, supra note 26, at 942.

85 See Fitzgerald & Jacobson, supra note 74, at 4.

86 See Trinka et al., supra note 11, at 2129 (stating that the FDA claims there is no evidence against the interchangeability of generics and brand name drugs while 81.6% of neurologists do not consider the FDA standards for AEDs sufficient).

87 Sander et al., supra note 82, at 1887.

88 See id.

89 See id.; Gagne et al., supra note 52, at 349.

90 Sander et al., supra note 82, at 1894 (explaining that patients may be confused by the change in product appearance leading to discontinuity in treatment). Note also that such changes may represent significant challenges in biologic drug, “biosimilar,” or follow on biologic formulations, including immunogenicity reactions. See, e.g., Liang, Bryan A. & Mackey, Tim, Emerging Patient Safety Issues Under Healthcare Reform: Follow On Biologics and Immunogenicity, 7 THERAPEUTICS & CLIN. RISK MANAG. 483 (2011).Google ScholarPubMed

91 Sander et al., supra note 82, at 1889 (stating that the FDA requires packaging and labeling to be similar to avoid confusion); see also Krauss et al., supra note 44, at 225 (observing in 17% of generic-to-generic switches a greater than 15% difference in AUC, and in 39% of generic-to-generic switches a greater than 15% difference in Cmax).

92 See Trinka et al., supra note 11, at 2129.

93 See Duh et al., supra note 81, at 2317.

94 Sankar et al., supra note 66, at 117, 121.

95 Differences in excipients, packaging, and labeling can be expected but not relevant to therapeutic quality. This may cause patient confusion. The formulation stability may also slightly vary. See Sander et al., supra note 82, at 1889. Several other European organizations from Germany, Italy, and Sweden recommend not switching from brand-named AEDs to generics. See Trinka et al., supra note 11, at 2129. In Pliva, Inc. v. Mensing, the Court held that generic manufacturers are unable to independently change their labels, even if dangerous side effects are known. Pliva, Inc. v. Mensing, 131 S. Ct. 2567, 2570 (2011). Generic drugs must have identical labels to brand name drugs according to federal law. See id. at 2572 (holding that federal drug regulations regarding generic drug manufacturers directly conflict with state laws and thus pre-empt the state claims); see also 21 U.S.C. § 355(j)(4)(g) (2006).

96 Duh et al., supra note 81, at 2318 (showing a chart of several societies’ views on AED generic substitution).

97 See Yamada & Welty, supra note 8, at 1408.

98 Sankar et al., supra note 66, at 115, 117.

99 See id. at 119 (showing a table of poor outcomes due to generic substitution).

100 Id. at 118 (explaining lamotrigine is a seizure medication).

101 Id.

102 Upon further analysis, 50% of the cases involved loss of seizure control, and most patients who switched back to the branded product were able to regain control. Id. In another study with 1354 patients, AED switchback rates were around 12.9% compared to other agents used to treat chronic disease such as statins and selective serotonin reuptake inhibitors, which had switchback rates ranging from 1.5% to 2.9%. See id.; see also Trinka et al., supra note 11, at 2130 (finding that switchback rates for other AEDs, such as clobazam and valproic acid, to branded drugs were around 20%).

103 Sankar et al., supra note 66, at 118; see, e.g., Bialer & Midha, supra note 26, at 943 (showing that the Danish National Drug Agency narrowed the BE 90% confidence interval limits to 90% to 111% primarily due to lamotrigine).

104 Duh et al., supra note 81, at 2317.

105 See Sankar et al., supra note 66, at 118.

106 Trinka et al., supra note 11, at 2130 (explaining that average BE addresses the issue of prescribability but does not address switchability, as switchability requires individual BE studies rather than the current average BE approach).

107 See Fitzgerald & Jacobson, supra note 74, at 4 (showing four case studies as examples of loss of seizure control when patients switched from brand-name drug to generic Levetiracetam).

108 A recent case-controlled study found that people with epilepsy who switched to a generic within the last six months had an 81% greater probability of receiving ambulance, emergency room, or inpatient care. See Sankar et al., supra note 66, at 118, 120. In this study, the generics were A-Rated AEDs, which according to the FDA are bioequivalent and interchangeable. Id. at 117-18.

109 Ghate, Sameer R. et al., Hemorrhagic and Thrombotic Events Associated with Generic Substitution of Warfarin in Paitients with Atrial Firbrillation: A Retrospective Analysis, 45 ANNALS PHARMACOTHERAPY 701, 701 (2011).CrossRefGoogle Scholar

110 See id. at 702.

111 See id. at 701.

112 The Thomas Reuters Medstat Market Scan Commercial Claims & Encounters and Medicare Supplimental & Coordination of Benefits database was analyzed to create a large retrospective study that included more than 37,750 patients regarding the clinical impact of warfarin product substitutions. See id. at 705. The conclusion of the study emphasized that patients should remain on the same product and avoid switching. See id. at 707. The authors found a 1.5- to 2.0-fold increase in risk of bleeding and thrombotic complications when switching products either from brand to generic, generic to brand, or generic to generic. See id. at 708.

113 See id. at 701.

114 See Dentali et al., supra note 80, at 386, 393.

115 See id. at 391.

116 In a study of warfarin and Coumadin, the generic showed small but significant decreases in INR, but no differences in morbidity or mortality. See Kesselheim, Aaron S. et al., Clinical Equivalence of Generic and Brand-Name Drugs Used in Cardiovascular Disease, A Systemic Review and Meta-Analysis, 300 JAMA 2514, 2523-24 (2008)CrossRefGoogle Scholar. Increased risk potential should be monitored more closely. Id. The authors found the differences generally to be insignificant. See id.; see also Todd Neale, Generic Cardiovascular Drugs Measure Up to Brand-Name Counterparts, MEDPAGE TODAY (Dec. 2, 2008), http://www.medpagetoday.com/Cardiology/MyocardialInfarction/11984. A citizens’ petition can be filed if a drug raises questions of quality. See Citizen Petition, GENERIC PHARMACEUTICAL ASSOC., http://www.gphaonline.org/issues/citizen-petitions (last visited July 8, 2012). Brand name drug companies have a long history of filing Citizen Petitions in an attempt to prolong the time in which generic drugs can enter the market. Id. Nearly three quarters of petitions in 2008 were found to have no merit and were consequently dismissed. Id. In response, the FDA Revitalization Act of 2007 aims to curb such abuses by forcing the FDA to respond to petitions within six months. Id. This is costly, but enables generic drugs to enter the market with less delay.

117 Ghate et al., supra note 109, at 709-10. Studies from the 1980s at Boston City Hospital indicated that after a routine substitution of Panwarfarin in place of brand name Coumadin, a number of patients began to experience wide fluctuation in their prothrombin times. See Haines, supra note 8, at 807. In response to this issue, the hospital did a retrospective study and found that 93% of patients that received Panwarfarin had a prothrombin time outside their target range. Id. Patients that switched were twice as likely to have prothrombin target times above the goal. Id. Also, patients that switched cost more due to increased clinical visits. Id. The estimated cost was $14,000 more than the non-switching group, which far exceeds any cost savings gained by switching to generic. Id. However, this study was done in the 1980s and involved fifteen patients who switched to the generic. This study is neither large nor recent enough to give current accurate information. Through the 1990s, as more generic warfarin products became available, pharmacy literature began to report suspicions that generics were not bioequivalent despite meeting FDA standards. Id. at 808. A 210-patient study by the Group Health Cooperative of Puget Sound found that patients who voluntarily switched to generics found no increased likelihood of increased change in INR. Id. Another study by the Barnes-Jewish Hospital Blood Thinner Clinic supported the finding that no differences between generic and brand named exist. Id. Kaiser examined 2300 patient records and found no evidence of failed generic warfarin—the only known prospective, randomized, controlled, and blinded clinical study failed to find substantial differences in patient response to a generic switch. Id. Several studies have proven that generic and brand warfarin are the same, but all the randomized controlled trials were very small, making them inadequate in detecting differences in products. Id. These retrospective studies may also be important for what they did not measure. For example, over-the-counter drugs that may affect the risk of hemorrhagic and thrombotic events are not captured in the database of warfarin users. See Carter, Barry L., Letter to the Editor, Equivalence of Generic and Brand-Name Drugs for Cardiovascular Disease, 301 JAMA 1654 (2009)CrossRefGoogle ScholarPubMed (finding that many of the systemic reviews do not consider the effects of repeated formulation substitution over time among a branded drug and its generics).

118 See Kesselheim et al., supra note 116.

119 Five prospective studies indicated stable INR values during a cross-over study entailing switching of drugs. See Haines, supra note 8, at 808. The issue with this study is that all five studies were too small. Id. A potentially “clinically important difference between the tested products might not have been detected due to inadequate sample size.” Id. In total, eleven studies were included for analysis. Dentali et al., supra note 80, at 386. Five were randomized controlled studies and six were observational, totaling more than 40,000 patients in all. Id. Sources used to search for warfarin studies included “MEDLINE and EMBASE, electronic records of meetings’ abstracts, and reference lists.” Id. Data used was required to be original studies, which enrolled patients using anticoagulant treatments and compared approved generic warfarin with brand name warfarin in either a clinical, laboratory, or management outcome. Id. In cross-over studies, no statistically significant differences were found among patients that switched from brand products to generics. Id. The results for observational studies also generally indicated little difference from brand name to generic drugs; however, in one study, change in therapeutic INR control after the switch to generic warfarin was reported at the individual patient level. Id. In this study, 72% of patients experienced a 10% or greater change in therapeutic INR control after the switch to generic warfarin. Id. at 391. Of these patients, about half experienced an improvement in INR control, and half experienced a worsening of INR control. Id. The conflicting information is a result of several different types of studies, all of which have significant limitations. Id. at 393. The authors noted that all the included studies have small sample sizes, and thus the analysis lacks power to detect even large differences in clinical event rate. Id. Three of the crossover RCTs that measured mean INR differences between brand and generic warfarin reported data as pooled averages, which fails to reflect individual responses of patients. Id. at 391. This is problematic given that substantial changes in individual INR control can occur despite minimal differences in mean INR response in the overall population.

120 Bupropion is the active ingredient of Wellbutrin, a commonly prescribed antidepressant. Several pharmaceutical alternatives have been developed, as well as generics for each alternative. See Chen, Mei-Ling et al., Challenges and Opportunities in Establishing Scientific and Regulatory Standards for Assuring Therapeutic Equivalence of Modified-Release Products: Workshop Summary Report, 40 EUR. J. PHARMACEUTICAL SCI. 148, 149 (2010)CrossRefGoogle Scholar; see also Howland, Robert H., When Is a “Generic” Medication Not Really a Generic?, 27 J. PSYCHOSOCIAL NURSING & MENTAL HEALTH SERVICES 13 (2010).Google Scholar

121 Review of Therapeutic Equivalence Generic Bupropion XL 300 mg and Wellbutrin XL 300 mg, supra note 80; see also Jacqueline Stenson, Report Questions Generic Antidepressant, NBC NEWS (Oct. 12, 2007), www.msnbc.msn.com/id/21142869/#.trgv2bzfk29.

122 Review of Therapeutic Equivalence Generic Bupropion XL 300 mg and Wellbutrin XL 300 mg, supra note 80 (explaining that higher doses of bupropion are associated with increased risk of seizure, so the 300 mg strength was not studied).

123 Id.; see also Stenson, supra note 121.

124 Chen et al., supra note 120, at 148-53.

125 Review of Therapeutic Equivalence Generic Bupropion XL 300 mg and Wellbutrin XL 300 mg, supra note 80.

126 See Chen et al., supra note 120, at 152 (finding that it is unclear if the efficacy of the drug is dependent on its peak or steady state concentrations of bupropion and/or its metabolites, but a therapeutic effect is usually observed after one to two weeks).

127 Generic Drug Equality Questioned: The People's Pharmacy, THE PEOPLE's PHARMACY (Oct. 12, 2007), http://www.peoplespharmacy.com/2007/10/12/generic-drug-eq/.

128 Stenson, supra note 121.

129 Id.

130 Id.; see also Drug Tests: Wellbutrin vs. Generic Bupropion, CONSUMERLAB, www.consumerlab.com/howtested/wellbutrin_vs_Generic_Bupropion/Wellbutrin/ (last visited Oct. 17, 2012).

131 See Stenson, supra note 121.

132 Review of Therapeutic Equivalence Generic Bupropion XL 300 mg and Wellbutrin XL 300 mg, supra note 80 (finding that if 10,000 persons switched from the brand name product to the generic product, then 500 to 800 people would be expected to experience a worsening of symptoms even if there was no change in drugs, as this is the natural course of the disease).

133 See id.

134 See mm, Comment to Generic Drug Equality Questioned, supra note 127 (Oct. 12, 2007, 11:14 AM) (“I too take Wellbutrin XL 300 mg and was switched from brand name to Teva's generic version by my pharmacy/health plan. I experienced a huge spike in depression that had been well controlled on brand name Wellbutrin for several years previously. Normally I have no problems taking generics, but with these findings, and my increase in depression this last year, I have to question the efficacy of Teva's version of Wellbutrin XL.”); see also Melissa, Comment to Generic Drug Equality Questioned, supra note 127 (Oct, 12, 2007 1:48 PM) (“I thought I might be imagining things when I switched from name brand Wellbutrin to generic Wellbutrin and began having depression symptoms. I was on 300 mg daily and though [sic] that maybe it was time to up the dosage or switch to another antidepressant when it didn't seem to work anymore. I suspected the generic drug might be the problem so a [sic] asked for a month's supply of the name brand and felt better almost instantly. I’m glad to know that this issue is being looked at.”).

135 See Melissa, supra note 134.

136 Review of Therapeutic Equivalence Generic Bupropion XL 300 mg and Wellbutrin XL 300 mg, supra note 80.

137 David Maris, A Drug Recall That Should Frighten Us All About the FDA, FORBES (Oct. 10, 2012, 3:00 PM), http://www.forbes.com/sites/davidmaris/2012/10/10/fda-recall-points-to-serious-problems-at-the-fda/.

138 Id.

139 Id.

140 Id.

141 Id.; see also Daniel J. DeNoon, Teva's High-Dose Generic Wellbutrin XL Withdrawn, WEBMD (Oct. 5, 2012), http://www.webmd.com/depression/news/20121005/teva-high-dose-generic-wellbutrin-xl-withdrawn.

142 A higher prevalence of side effects caused three patients to drop out of the study on the first day. Increased side effects could lead to poorer compliance of medical care. See Chenu, Franck et al., Comparison of Pharmacokinetic Profiles of Brand-Name and Generic Formulations of Citalopram and Venlafaxine: A Crossover Study, 70 J. CLINICAL PSYCHIATRY 958, 965 (2009)CrossRefGoogle ScholarPubMed. Personal and familial suffering resulting from potential relapse may not be worth the immediate savings of generics. Id.

143 See Sankar et al., supra note 66, at 120 (showing a graph of states and their respective policy for substitution); see also Therapeutic Research Ctr., State Regulations on Generic Substitution, PHARMACIST's LETTER (2006), http://pharmacistsletter.therapeuticresearch.com/pl/ArticleDD.aspx?nidchk=1&cs=&s=PL&pt=2&segment=1186&dd=220901 (last updated Apr. 2009) (presenting a table of each state's policy).

144 Trinka et al., supra note 11, at 2131.

145 Ghate et al., supra note 109, at 706.

146 The mini-Sentinel pilot program is currently underway to build a nationwide electronic safety surveillance system for drugs and medical products. See MINI-SENTINEL, http://mini-sentinel.org/ (last visited Sept. 24, 2012) (showing reports of pilot projects and progress). This system aims to “help scientists better understand the potential safety issues associated with FDA-approved products.” FDA's “Mini-Sentinel” Safety Pilot Program is Up and Running, Demonstrating Rapid Analysis of Medical Product Safety Questions, FDA, http://www.fda.gov/downloads/Safety/FDAsSentinelInitiative/UCM268035.pdf (last visited Sept. 20, 2012). This system may offer great data and insight into issues associated with BE in the future.

147 Trinka et al., supra note 11, at 2130-31 (showing that after generic substitution there were higher utilization rates of medical services and longer hospital stays).

148 More event reporting needs to take place as under reporting may be an issue. Physicians may be unaware of a switch to a generic drug, which may also inhibit studies evaluating the effects of switching from branded to generic drugs. Reporting events to Med Watch is an important step to better tracking issues relating to drug switches. See Fitzgerald & Jacobson, supra note 74, at 4.

149 See Trinka et al., supra note 11, at 2129 (“For decades the establishment of bioequivalence has generally relied on the comparison of population averages … .”).

150 See, e.g., Bialer & Midha, supra note 26, at 945 (explaining that the traditional average BE approach gives no information about patient variances associated with generic and brand name drugs).

151 See CDER, supra note 25, at 24 (showing statistical equations used in BE criteria).

152 Trinka et al, supra note 11, at 3; see also Welage et al., supra note 7, at 864.

153 See Bialer & Midha, supra note 26, at 945 (explaining that the scaled average BE approach first evaluates the average variance of the reference formulation, then sets that range for the confidence interval for the test product according to the reference variance).

154 James V. Hennessey, Presentation at the Meeting of the Advisory Committee for Pharmaceutical Science and Clinical Pharmacology (July 26, 2011), available at https://www.aace.com/files/views/leg_reg_hennessey_july_2011.pdf.

155 See id.; see also Bialer & Midha, supra note 26, at 946.

156 Hennessey, supra note 154.

157 See Karalis, Vengelis et al., On the Leveling-off Properties of the New Bioequivalence Limits for Highly Variable Drugs of the EMA Guideline, 44 EUR. J. PHARMACEUTICAL SCI. 497, 498 (2011)CrossRefGoogle Scholar; see also Hennessey, supra note 154.

158 See GARY BEUHLER, FDA, HISTORY OF BIOEQUIVALENCE FOR CRITICAL DOSE DRUGS 69 (2010), available at http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/AdvisoryCommitteeforPharmaceuticalScienceandClinicalPharmacology/UCM209319.pdf.

159 See WAPLES, YVETTE & TOPP, ELIZABETH M., FDA, SUMMARY MINUTES OF THE ADVISORY COMMITTEE FOR PHARMACEUTICAL SCIENCE AND CLINICAL PHARMACOLOGY 1, 5 (2011), available at http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/AdvisoryCommitteeforPharmaceuticalScienceandClinicalPharmacology/UCM272111.pdf.Google Scholar

160 Krauss et al., supra note 44, at 227.

161 Welage et al., supra note 7, at 854.

162 See FDA Considering Tighter Controls on Generics, supra note 51 (reporting that in an April 2010 meeting, the FDA advisory committee judged that current standards were not sufficient for critical dose drugs); Kristina Fiore, FDA May Rein in Bioequivalence for Critical Dose Drugs, MEDPAGE TODAY (Oct. 23, 2010), http://www.medpagetoday.com/PublicHealthPolicy/PublicHealth/22930; see also BEUHLER, supra note 158 (showing that the FDA discussed a need for narrowing the BE range).

163 See WAPLES & TOPP, supra note 159, at 5.

164 See Yamada & Welty, supra note 8, at 1413.

165 EUR. MED. AGENCY, GUIDELINE ON THE INVESTIGATION OF BIOEQUIVALENCE 15 (2008), available at http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2009/09/WC500003011.pdf (acknowledging that NTI drugs may need a tighter interval and suggests narrowing the BE range to 90 to 111% in special cases of NTI drugs).

166 See BEUHLER, supra note 158.

167 The U.S. version of Tegretol is different from the European version, and thus the generics made from each reference drug will be different. Creating an international reference sample standard for each drug will make generics safer. See Bialer & Midha, supra note 26, at 945.

168 See Arnold, supra note 64, at 257.

169 See id.

170 See Bapuji et al., supra note 3, at 101.

171 See EUR. MED. AGENCY, supra note 165, at 15.

172 See generally AlAmeri et al., supra note 44, at 691-95.

173 Duh et al., supra note 81, at 2320.

174 See AlAmeri et al., supra note 44, at 692.