Abstract
Importance
Some practice guidelines warn against generic L-thyroxine preparation switching.
Objective
To examine the rates of generic L-thyroxine preparation switching within one year of initiating L-thyroxine, and to examine factors associated with switching.
Design and setting
Retrospective study using national data from a large administrative claims database from January 2008 through November 2018.
Patients
Medicare or commercially insured adults (≥18 years) who filled a generic L-thyroxine preparation.
Main outcome measures
At least one switch from one generic L-thyroxine preparation to another within 1 year of L-thyroxine initiation defined by prescription fills.
Results
From January 2008 to November 2018, we included 483,390 patients who initiated generic L-thyroxine: mean (SD) age was 61.4 years (15.2), 75.2% were female, 72.6% were white. Within 1 year of initiating therapy, 98,013 (20%) switched to another L-thyroxine generic preparation at least once. In a multivariate logistic regression analysis, factors associated with switching included the number of pharmacies visited to fill L-thyroxine (>2 vs 1 adjusted OR [aOR] 7.15, 95% confidence interval [CI] 6.97–7.34), age ≥75 vs. <45 years (aOR 1.29, 95% CI 1.26–1.33), history of thyroid surgery (aOR 1.22, 95% CI 1.13–1.31), and first L-thyroxine fill date in 2018 vs. 2008 (aOR 3.32, 95% CI 3.14–3.51).
Conclusions and relevance
One in five patients switched among generic L-thyroxine manufacturers within one year of treatment initiation. Generic L-thyroxine switching occurred more often when more pharmacies were used to fill L-thyroxine. Given existing guideline recommendations, additional studies should clarify the impact of generic L-thyroxine switching on thyroid hormone values.
Keywords: Generic, Levothyroxine, Hypothyroidism
Introduction
L-thyroxine is one of the most commonly prescribed drugs in the United States. L-thyroxine is available as a generic or brand-name product [1, 2]. In 2018, there were about 96 million levothyroxine prescriptions in the United States. It is estimated that between 80–85% of these prescriptions were for generic-name L-thyroxine. Over the last decade, there has been an increase in the rate of generic L-thyroxine dispensing, from 60% of all fills in 2007 to 84.9% in 2016. Primary care vs. endocrine providers are more likely to prescribe generic vs. brand L-thyroxine in the U.S. [2]. The 2014 American Thyroid Association Guideline recognizes that there is no evidence regarding the superiority of one particular L-thyroxine product over another and recommends that clinicians choose between brand L-thyroxine or, alternatively, maintenance of the same generic preparation [3]. Importantly, the guideline highlights the need to maintain the same generic preparation due to concerns that even products judged to be bioequivalent may not have therapeutic equivalence, and that switching could lead to changes in serum TSH. If switching were to occur, for instance, from one generic L-thyroxine manufacturer to another, the guideline further recommends (weak recommendation based on low-quality evidence) that thyroid-stimulating hormone (TSH) laboratory values should be monitored post switch to ensure that levels remain consistent with appropriate treatment [3]. However, the implementation of this recommendation for patients taking generic L-thyroxine is challenging as switching between generic L-thyroxine preparations may occur at the level of the dispensing pharmacy, without the patient’s or clinician’s awareness. Consequently, clinicians cannot order recommended laboratory testing. Furthermore, this guideline recommendation is at odds with the Food and Drug Administration (FDA) position that approved L-generic products with the expectation that they are substitutable without the need for TSH testing [4, 5].
Given the rising numbers of patients receiving generic L-thyroxine, and the contradicting recommendations about switching between L-thyroxine products, it is important to understand how often switching among generic L-thyroxine preparations occurs. We used a national administrative claims database to examine the rates of generic L-thyroxine preparation switching within one year of initiating L-thyroxine, as well as the factors associated with switching.
Methods
Study design and data source
We conducted a retrospective study of de-identified administrative claims data linked with laboratory results from a large database, OptumLabs™ Data Warehouse (OLDW), which includes commercially insured and Medicare Advantage enrollees throughout the United States [6]. The database contains longitudinal health information on enrollees representing a diverse mix of ages, ethnicities. The dataset used in this analysis includes many insured patients in the United States, with the greatest representation from the South and Midwest US census regions. However, this predominance of specific geographic areas matches the representation of those areas to the general US population. [7]. Furthermore, our included population reflects the pattern of prescription by specialty found previously, with the majority of generic L-thyroxine prescribed by primary care providers and less frequently by endocrinologists [2]. The health plan provides comprehensive full insurance coverage for physician, hospital, and prescription drug services. Pharmacy claims include information on medications dispensed, including doses; amount dispensed; and dates of prescriptions. Study data were accessed using techniques compliant with the Health Insurance Portability and Accountability Act of 1996. Because this study involved analysis of pre-existing, de-identified data, the Mayo Clinic Institutional Review Board declared it exempt from board approval.
Study population
We included adults (≥18 years) who newly initiated generic L-thyroxine preparations from any of the three most common manufacturers (Mylan, Sandoz, or Lannett) between January 1st, 2008 and November 30th, 2018. Patients were required not to have filled brand L-thyroxine before initiation of generic L-thyroxine. In the United States, L-thyroxine is dispensed in medication bottlers containing the number of pills prescribed by clinicians per each refill; as such, we have the information on the day’s supply for each pill. Using that information, patients were required to have at least 12 months of continuous medical and pharmacy benefits after the L-thyroxine fill and had to be on L-thyroxine treatment for at least 12 months after initiation. Discontinuation of medication was defined as not refilling a prescription within 40 days after the end of supply of the last filled prescription. During the 12 months after L-thyroxine treatment initiation, we defined index date of first switch when patients filled a different L-thyroxine preparation from the original fill. For patients who continued to fill the same sourced L-thyroxine, the index date was a random fill date within one year of initiation. Patients were required to have continuous medical and pharmacy benefits for at least 6 months before the index date. This ensured that at least 6 months of information on patients’ past medical history and medication use were available for purposes of identifying baseline characteristics.
We excluded adults who switched to brand-name L-thyroxine at any point within the first year of initiation of L-thyroxine therapy, were pregnant, had diagnosed hypopituitarism or hyperthyroidism (diagnostic codes in Appendix Table 3), and those who filled other forms of thyroid replacement therapy before index date, including thyroid extracts, or T3 therapy, such as liothyronine, thyroid desiccated/extracts; as well as the following brands: Cytomel, Armour Thyroid, or Nature Thyroid (Study flow chart in Appendix Fig. 2).
Baseline characteristics
Baseline patient characteristics included age, sex, race/ethnicity, census region, provider specialty (general provider vs. endocrinology vs. other specialist), L-thyroxine initial dose (calculated based on fill data), year of L-thyroxine initial fill, length of L-thyroxine use before switching; number of L-thyroxine fills; number of pharmacies visited to fill L-thyroxine; health plan type (commercial vs. Medicare Advantage), Charlson score (estimated using ICD-9-CM/ICD-10 diagnosis codes included in administrative claims) [8, 9]; baseline comorbidities (atrial fibrillation, myocardial infarction, congestive heart failure, stroke, thyroid cancer, and history of previous thyroid surgery). We hypothesized that these comorbidities may be associated with more frequent pharmacy visits and thus, higher rates of switching.
Exposures
We used First Databank to categorize fills as generic and to identify the specific L-thyroxine manufacturer. First Databank categorizes pharmacy products as generic if they are sold under a generic pharmacy label. The L-thyroxine generic manufacturers included Mylan, Sandoz, and Lannett. These three generic products comprised 99% of all the generic L-thyroxine fills.
Study outcome
Switching occurred when patients starting with a generic L-thyroxine product made by any of the 3 major manufacturers switched to a preparation made by any other generic manufacturer, including other small manufactures for generic L-thyroxine. If patients subsequently switched back to the original manufacturer, we also counted this as a switch. Switchers were then classified as those who switched once, twice, and more than 2 times within one year of index date. Patients who did not switch were classified as non-switchers.
Statistical analyses
Baseline characteristics analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC) and Stata version 14.1 (StatCorp, College Station, TX). Baseline characteristics of non-switchers and switchers were reported, including those who switched once, twice, and more than two times within one year after initiating generic L-thyroxine. Categorical variables were reported as frequencies with percentages, whereas continuous variables were reported as means with standard deviations (SDs), median, Q1 and Q3. Logistic regression was used to assess factors associated with switching compared to non-switching adjusted for all the variables listed in Table 1. Results were presented as adjusted odds ratios (aORs) and 95% confidence intervals (CIs). Logistic regression analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC) and R 4.0.2 (R Foundation).
Table 1.
Descriptive factors for patients who initiated generic L-thyroxine and continued on the same generic preparation (non-switchers), or switched to another generic L-thyroxine preparation (switchers), including those who subsequently switched once, twice, or more than twice within one year of L-thyroxine initiation
| Non-switcher (N = 385,377) | Switcher |
Total (N = 483,390) | ||||
|---|---|---|---|---|---|---|
| All switchers (N = 98,013) | Switch once (N = 66,161) | Switch twice (N = 23,651) | Switch > 2 (N = 8201) | |||
|
| ||||||
| Age | ||||||
| Mean (SD) | 61.4 (15.2) | 61.7 (15.3) | 62.2 (15.0) | 61.2 (15.6) | 59.0 (16.1) | 61.4 (15.2) |
| Median | 64.0 | 64.0 | 65.0 | 64.0 | 60.0 | 64.0 |
| Q1, Q3 | 51.0, 73.0 | 52.0, 73.0 | 53.0, 73.0 | 51.0, 73.0 | 48.0, 71.0 | 51.0, 73.0 |
| Age group | ||||||
| 18–44 | 59,225 (15.4%) | 14,381 (14.7%) | 9011 (13.6%) | 3789 (16.0%) | 1581 (19.3%) | 73,606 (15.2%) |
| 45–54 | 60,261 (15.6%) | 14,702 (15.0%) | 9693 (14.7%) | 3561 (15.1%) | 1448 (17.7%) | 74,963 (15.5%) |
| 55–64 | 77,145 (20.0%) | 20,454 (20.9%) | 13,718 (20.7%) | 4828 (20.4%) | 1908 (23.3%) | 97,599 (20.2%) |
| 65–74 | 105,499 (27.4%) | 27,078 (27.6%) | 18,948 (28.6%) | 6401 (27.1%) | 1729 (21.1%) | 132,577 (27.4%) |
| 75+ | 83,247 (21.6%) | 21,398 (21.8%) | 14,791 (22.4%) | 5072 (21.4%) | 1535 (18.7%) | 104,645 (21.6%) |
| Gender | ||||||
| Female | 290,254 (75.3%) | 73,256 (74.7%) | 49,518 (74.8%) | 17,725 (74.9%) | 6013 (73.3%) | 363,510 (75.2%) |
| Male | 95,123 (24.7%) | 24,757 (25.3%) | 16,643 (25.2%) | 5926 (25.1%) | 2188 (26.7%) | 119,880 (24.8%) |
| Race | ||||||
| Asian | 12,417 (3.2%) | 3313 (3.4%) | 2265 (3.4%) | 787 (3.3%) | 261 (3.2%) | 15,730 (3.3%) |
| Black | 35,516 (9.2%) | 8696 (8.9%) | 5894 (8.9%) | 2060 (8.7%) | 742 (9.0%) | 44,212 (9.1%) |
| Hispanic | 35,262 (9.2%) | 8363 (8.5%) | 5441 (8.2%) | 2151 (9.1%) | 771 (9.4%) | 43,625 (9.0%) |
| White | 280,604 (72.8%) | 70,301 (71.7%) | 47,939 (72.5%) | 16,588 (70.1%) | 5774 (70.4%) | 350,905 (72.6%) |
| Unknown | 21,578 (5.6%) | 7340 (7.5%) | 4622 (7.0%) | 2065 (8.7%) | 653 (8.0%) | 28,918 (6.0%) |
| L-T4 dose (initial) | ||||||
| ≤50 MCG | 171,081 (44.4%) | 43,341 (44.2%) | 28,772 (43.5%) | 10,468 (44.3%) | 4101 (50.0%) | 214,422 (44.4%) |
| 51–100 MCG | 129,859 (33.7%) | 31,260 (31.9%) | 21,777 (32.9%) | 7281 (30.8%) | 2202 (26.9%) | 161,119 (33.3%) |
| 101–200 MCG | 84,002 (21.8%) | 23,294 (23.8%) | 15,538 (23.5%) | 5876 (24.8%) | 1880 (22.9%) | 107,296 (22.2%) |
| >200 MCG | 435 (0.1%) | 118 (0.1%) | 74 (0.1%) | 26 (0.1%) | 18 (0.2%) | 553 (0.1%) |
| Length of L-T4 before index date | ||||||
| Mean (SD) | 215.1 (97.6) | 204.4 (96.2) | 221.1 (97.1) | 182.4 (82.4) | 133.0 (79.2) | 212.9 (97.4) |
| Median | 224.0 | 207.0 | 231.0 | 188.0 | 128.0 | 219.0 |
| Q1, Q3 | 160.0, 291.0 | 129.0, 281.0 | 156.0, 301.0 | 118.0, 248.0 | 62.0, 192.0 | 155.0, 289.0 |
| No. of T4 fills within 1 year | ||||||
| Mean (SD) | 8.9 (3.8) | 8.9 (3.7) | 8.6 (3.7) | 9.3 (3.7) | 10.7 (3.7) | 8.9 (3.8) |
| Median | 10.0 | 9.0 | 8.0 | 10.0 | 11.0 | 10.0 |
| Q1, Q3 | 5.0, 12.0 | 5.0, 12.0 | 5.0, 12.0 | 6.0, 12.0 | 8.0, 13.0 | 5.0, 12.0 |
| No. of T4 fills within 1 year | ||||||
| 1–5 times | 127,920 (33.2%) | 26,897 (27.4%) | 20,776 (31.4%) | 5545 (23.4%) | 576 (7.0%) | 154,817 (32.0%) |
| 6–10 times | 82,100 (21.3%) | 31,759 (32.4%) | 21,475 (32.5%) | 7276 (30.8%) | 3008 (36.7%) | 113,859 (23.6%) |
| ≥11 times | 175,357 (45.5%) | 39,357 (40.2%) | 23,910 (36.1%) | 10,830 (45.8%) | 4617 (56.3%) | 214,714 (44.4%) |
| No. of pharmacies visited to fill L-T4 | ||||||
| Mean (SD) | 1.3 (0.6) | 1.8 (0.8) | 1.8 (0.8) | 1.7 (0.8) | 1.9 (1.0) | 1.4 (0.7) |
| Median | 1.0 | 2.0 | 2.0 | 1.0 | 2.0 | 1.0 |
| Q1, Q3 | 1.0, 1.0 | 1.0, 2.0 | 1.0, 2.0 | 1.0, 2.0 | 1.0, 2.0 | 1.0, 2.0 |
| Health plan | ||||||
| Commercial | 182,001 (47.2%) | 43,968 (44.9%) | 28,741 (43.4%) | 10817 (45.7%) | 4410 (53.8%) | 225,969 (46.7%) |
| Medicare advantage | 203,376 (52.8%) | 54,045 (55.1%) | 37,420 (56.6%) | 12834 (54.3%) | 3791 (46.2%) | 257,421 (53.3%) |
| Census region | ||||||
| Midwest | 106,697 (27.7%) | 28,188 (28.8%) | 18,818 (28.4%) | 6987 (29.5%) | 2383 (29.1%) | 134,885 (27.9%) |
| Northeast | 53,699 (13.9%) | 13,190 (13.5%) | 9018 (13.6%) | 3074 (13.0%) | 1098 (13.4%) | 66,889 (13.8%) |
| South | 168,652 (43.8%) | 41,976 (42.8%) | 28,595 (43.2%) | 9977 (42.2%) | 3404 (41.5%) | 210,628 (43.6%) |
| West | > 56,320 (>14%)a | >14,650 (>14%)a | >9720 (>14%)a | 3613 (15.3%) | 1316 (16.0%) | >70,980 (>14%)a |
| Other | <11a | <11a | <11a | 0 | 0 | <11a |
| Prescribing provider | ||||||
| Endocrinology | 24,771 (6.4%) | 6018 (6.1%) | 3850 (5.8%) | 1589 (6.7%) | 579 (7.1%) | 30,789 (6.4%) |
| Primary care | 280,750 (72.9%) | 68,189 (69.6%) | 46,664 (70.5%) | 16,200 (68.5%) | 5325 (64.9%) | 348,939 (72.2%) |
| Other | 44,435 (11.5%) | 12,467 (12.7%) | 8048 (12.2%) | 3062 (12.9%) | 1357 (16.5%) | 56,902 (11.8%) |
| Missing/Unknown | 35,421 (9.2%) | 11,339 (11.6%) | 7599 (11.5%) | 2800 (11.8%) | 940 (11.5%) | 46,760 (9.7%) |
| Charlson comorbidity index | ||||||
| Mean (SD) | 1.1 (1.7) | 1.2 (1.9) | 1.2 (1.8) | 1.2 (1.9) | 1.3 (2.1) | 1.1 (1.7) |
| Median | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| Q1, Q3 | 0.0, 1.0 | 0.0, 2.0 | 0.0, 2.0 | 0.0, 2.0 | 0.0, 2.0 | 0.0, 2.0 |
| Baseline comorbidities (within 6 months before index date) | ||||||
| Atrial fibrillation | 20,046 (5.2%) | 5765 (5.9%) | 3882 (5.9%) | 1412 (6.0%) | 471 (5.7%) | 25,811 (5.3%) |
| Myocardial infarction | 4551 (1.2%) | 1481 (1.5%) | 976 (1.5%) | 362 (1.5%) | 143 (1.7%) | 6032 (1.2%) |
| Congestive heart failure | 19,160 (5.0%) | 5778 (5.9%) | 3825 (5.8%) | 1447 (6.1%) | 506 (6.2%) | 24,938 (5.2%) |
| Stroke | 6620 (1.7%) | 2147 (2.2%) | 1447 (2.2%) | 512 (2.2%) | 188 (2.3%) | 8767 (1.8%) |
| Thyroid cancer | 4480 (1.2%) | 1622 (1.7%) | 953 (1.4%) | 431 (1.8%) | 238 (2.9%) | 6102 (1.3%) |
| Thyroid surgery (total) | 3149 (0.8%) | 1585 (1.6%) | 798 (1.2%) | 449 (1.9%) | 338 (4.1%) | 4734 (1.0%) |
| Year of index prescription | ||||||
| 2008 | 12,183 (3.2%) | 1952 (2.0%) | 1353 (2.0%) | 430 (1.8%) | 169 (2.1%) | 14,135 (2.9%) |
| 2009 | 23,692 (6.1%) | 3456 (3.5%) | 2582 (3.9%) | 633 (2.7%) | 241 (2.9%) | 27,148 (5.6%) |
| 2010 | 28,538 (7.4%) | 3938 (4.0%) | 2809 (4.2%) | 872 (3.7%) | 257 (3.1%) | 32,476 (6.7%) |
| 2011 | 31,235 (8.1%) | 5069 (5.2%) | 3451 (5.2%) | 1256 (5.3%) | 362 (4.4%) | 36,304 (7.5%) |
| 2012 | 32,710 (8.5%) | 5628 (5.7%) | 3923 (5.9%) | 1321 (5.6%) | 384 (4.7%) | 38,338 (7.9%) |
| 2013 | 31,294 (8.1%) | 7630 (7.8%) | 5381 (8.1%) | 1661 (7.0%) | 588 (7.2%) | 38,924 (8.1%) |
| 2014 | 35,529 (9.2%) | 13,076 (13.3%) | 9436 (14.3%) | 2699 (11.4%) | 941 (11.5%) | 48,605 (10.1%) |
| 2015 | 40,061 (10.4%) | 8381 (8.6%) | 5961 (9.0%) | 1762 (7.5%) | 658 (8.0%) | 48,442 (10.0%) |
| 2016 | 52,235 (13.6%) | 12,913 (13.2%) | 9182 (13.9%) | 2730 (11.5%) | 1001 (12.2%) | 65,148 (13.5%) |
| 2017 | 59,006 (15.3%) | 18,353 (18.7%) | 11,499 (17.4%) | 5075 (21.5%) | 1779 (21.7%) | 77,359 (16.0%) |
| 2018 | 38,894 (10.1%) | 17,617 (18.0%) | 10,584 (16.0%) | 5212 (22.0%) | 1821 (22.2%) | 56,511 (11.7%) |
N < 11 cannot be shared to protect patient confidentiality
Results
Four hundred eighty three thousand and three hundred and ninety (483,390) patients who initiated a generic L-thyroxine prescription between 1/1/2008 and 11/30/2018 were included in this study: mean (SD) age was 61.4 years (15.2), 75.2% were female, 72.6% were white, and 44.4% received a L-thyroxine dose ≤50 mcg per day. The mean (SD) number of pharmacies visited by patients to fill L-thyroxine was 1.4 (0.7), mean (SD) number of days on L-thyroxine before switching (index date for non-switchers) was 212.9 (97.4), and mean (SD) number of L-thyroxine fills within first year of therapy was 8.9 (3.8). At the initiation of treatment, the most common generic L-thyroxine manufacturer was Mylan (51%), followed by Lannet (37%), and Sandoz (12%).
Among all patients who started generic L-thyroxine, 385,377 (80%) did not switch their L-thyroxine preparation and 98,013 (20%) patients switched their L-thyroxine preparation at least once within one year of L-thyroxine initiation; 66,161 (14%) switched once, 23,651 (5%) switched twice, and 8201 (1%) switched more than 2 times (Table 1).
The logistic regression model used to examine the factors associated with switching had an area under the curve (AUC) of 0.75 (see Appendix Fig. 3 and Table 1). Multiple variables were significantly associated with generic L-thyroxine switching in the model, including L-thyroxine dose between 101 and 200 mcg/daily vs. ≤50 mcg/daily (OR 1.15; 95% CI 1.12, 1.17), number of pharmacies visited to fill L-thyroxine (2 pharmacies vs 1: OR 4.88; 95% CI 4.80, 4.97; >2 pharmacies vs 1: OR 7.15; 95% CI 6.97, 7.34), history of thyroid surgery (OR 1.22; 95% CI 1.13, 1.31), age between 45–54 years (OR 1.14; 95% CI 1.10, 1.17), age between 55–64 years (OR 1.20; 95% CI 1.17, 1,23), age between 65–74 year (OR 1.26; 95% CI 1.23, 1.30); and age older than 74 years (OR 1.29; 95%CI 1.26, 1.33), vs. age younger than 45 years. Furthermore, patients were more likely to experience generic L-thyroxine switching in 2014 (OR 2.62; 95% CI 2.48; 2.77), in 2017 (OR 2.22; 95% CI 2.10; 2.34) and 2018 (OR 3.32; 95% CI 3.14; 3.51) compared to 2008, as showed in Fig. 1, and Appendix Table 2.
Fig. 1.
Forest plot of factors associated with switching from a generic L-thyroxine preparation to another generic L-thyroxine preparation within one year of L-thyroxine initiation. The squares represent the point estimates of Log Odds Ratio of switchers vs non-switchers; the horizontal lines represent the 95% confidence interval. Any variable with a confidence interval not crossing the vertical line of the Log Odds Ratio of 0 represents a variable significantly associated with more (for positive values) or less (for negative values) generic L-thyroxine switching
Discussion
In this large retrospective longitudinal study set in the United States, using a national administrative claims database, we found that one in five patients who initiated generic L-thyroxine, and continued L-thyroxine treatment for at least 1 year, switched to a preparation made by another generic L-thyroxine manufacturer within one year of starting thyroid replacement therapy. Furthermore, we found that several factors were associated with higher rates of switching, including the number of pharmacies visited to fill the L-thyroxine prescription, patient’s age, L-thyroxine initiation during the more recent years, and history of thyroid surgery. Several studies have assessed the impact of brand L-thyroxine switching to another L-thyroxine formulation; however, to our knowledge, this is the first study that appraised the frequency of generic to generic L-thyroxine switching [10–12].
According to some of the current practice guidelines, the prevalence of switching from a specific generic L-thyroxine preparation to one made by a different manufacturer may have significant implications on patient care [3, 13]. The 2014 American Thyroid Association Guideline recommends followup testing upon switching between different L-thyroxine preparations in order to ensure stable TSH levels [3]. Since switching between generic L-thyroxine products made by different manufacturers occurs at the level of the pharmacy without clinicians’ express input or awareness, this recommendation is difficult to implement in clinical practice unless clinicians choose to prescribe a specific brand-name L-thyroxine product. This would ensure that patients receive the same manufacturer-sourced preparation upon every refill, as switching from brand-name L-thyroxine to another brand-name or to generic L-thyroxine would require notification of patients or clinicians. Although the use of brand-name L-thyroxine products might be preferred theoretically for these reasons, often this is not possible, because of cost considerations. Brand L-thyroxine is several times more expensive than generic L-thyroxine; information available from the OptumLabs Data Warehouse suggests that in 2016 the median out-of-pocket cost of a 30-day prescription for branded thyroid hormone was $25–$31 in 2016 across insured patients, substantially higher than the $2–$10 median out-of-pocket cost for a 30-day prescription for generic thyroid hormone products. Moreover, it has yet to be shown that switching among different generic manufacturers leads to variable TSH levels or affects clinical outcomes.
Based on our study, clinicians who choose to keep patients on generic L-thyroxine should recognize that there are some factors that may increase the probability of generic L-thyroxine switching. We found that the most important modifiable factor associated with switching was a greater number of pharmacies visited. More visits to different pharmacies to fill L-thyroxine may open opportunities to receive a L-thyroxine preparation sourced from a different manufacturer. Contrary, it is possible that visits to the same pharmacy may help patient retain the same preparation since the supply contractor is more likely to remain stable. Yet, many other factors, not measured in the study, may contribute to generic L-thyroxine manufacture switching even when patients visit the same pharmacy: changes between generic suppliers, or changes in contracts between manufacturers, insurers, and pharmacy benefits managers.
In our study, switching occurred more often during more recent years, in 2017 and 2018, compared to 2008. We hypothesize that drug shortages may fuel different L-thyroxine preparations dispensed. For instance, in 2017 and 2018, there was a significant shortage of L-thyroxine as a result of hurricanes Irma and Maria, which hit Puerto Rico and knocked out L-thyroxine manufacturing facilities [14]. Price increases may also be associated with high rates of switching. During 2014, a year associated with more generic L-thyroxine switching, there was a rise in the price of generic L-thyroxine and many other generic drugs. High generic prices may have affected the need for pharmacies to renegotiate contract agreements with drug suppliers, which, in turn, could have changed the L-thyroxine generic suppliers for many pharmacies [15].
We also found some patient and clinical factors associated with generic L-thyroxine switching. Older patients, patients receiving L-thyroxine after thyroid surgery, or those receiving doses higher than 100 mcg were more likely to experience switching. The reason behind this association is not clear. It is possible that these factors trigger a higher number of pharmacy visits due to more frequent L-thyroxine dose adjustments. For instance, patients with a history of thyroid surgery may need more frequent L-thyroxine dose adjustment, and thus more prescriptions, to reach target thyroid hormone levels [16]. The association between prescriptions for L-thyroxine and more pharmacy visits could still exist even though the number of prescriptions was not a significant factor associated with switching in our logistic regression model. Therefore, to minimize rates of switching clinicians may want to avoid frequent L-thyroxine prescriptions.
Our study has several limitations that need to be considered. The results of this study are based on United States data and are not generalizable to other countries with different regulatory policies for medication dispensing and generic substitution. We did not study the frequency of switching from brand to generic L-thyroxine (and vice versa), a relevant question for practice. Our rationale to exclude brand L-thyroxine switching is that brand L-switching to another L-thyroxine product is an event that clinicians and patients mostly direct in the US. As a result, when brand L-thyroxine switching occurs, the clinician could order post-switch TSH values as recommended by guidelines to ensure the levels remain consistent. Contrary, generic to generic L-thyroxine switching happens at the pharmacy level, and patients and clinicians do not control it and may not even be aware of it. Furthermore, these administrative data did not contain pharmacy-level information that could better elucidate other factors associated with L-thyroxine switching. For instance, we were not able to determine whether or not a pharmacy belongs to a particular chain (e.g., CVS). Moreover, due to the study’s retrospective nature and the use of claims data, possible misclassification of variables (e.g., medication fills) could have occurred. However, misclassification bias is present in all study designs. Indeed, using administrative data to evaluate prescription fills and type of medication fills is considered superior to self-reported medication used (susceptible to recall bias). Finally, our study sample includes mostly patients with commercial health insurance; the applicability of study findings to under-insured populations is unclear.
Conclusion
One in five patients initiating generic L-thyroxine switched to a preparation made by another generic L-thyroxine manufacturer within one year of treatment. While patients and clinicians are likely unaware of the switch among generic L-thyroxine manufacturers, several factors are associated with a higher rate of switching including the number of pharmacy visits. Additional studies should be conducted to clarify the impact of generic L-thyroxine switching on thyroid hormone values to inform guideline recommendations.
Funding
This project was supported by the Food and Drug Administration (FDA) of the U.S. Department of Health and Human Services (HHS) as part of a financial assistance award [Center of Excellence in Regulatory Science and Innovation grant to Yale University and Mayo Clinic, U01FD005938] totaling $158,430 with 100 percent funded by FDA/HHS. The contents are those of the author(s) and do not necessarily represent the official views of, nor an endorsement, by FDA/HHS, or the U.S. Government.
Appendix
Table 2.
Results of logistic regression for factors associated with switching from one generic L-thyroxine preparation to another
| Factors | OR | Lower | Upper | OR (95% CI) | p value |
|---|---|---|---|---|---|
|
| |||||
| Dose 51–100 mcg vs ≤50 mcg | 1.01 | 0.99 | 1.03 | 1.01 (0.99, 1.03) | 0.249 |
| Dose 101–200 mcg vs ≤50 mcg | 1.15 | 1.12 | 1.17 | 1.15 (1.12, 1.17) | <0.0001 |
| Dose >200 mcg vs ≤50 mcg | 1.05 | 0.84 | 1.31 | 1.05 (0.84, 1.31) | 0.6712 |
| Length of T4 before index date | 1.00 | 1.00 | 1.00 | 1.00 (1.00, 1.00) | <0.0001 |
| No. of T4 fills within 1 year: 6–10 vs 1–5 | 1.74 | 1.71 | 1.78 | 1.74 (1.71, 1.78) | <0.0001 |
| No. of T4 fills within 1 year: ≥11 vs 1–5 | 1.63 | 1.60 | 1.66 | 1.63 (1.60, 1.66) | <0.0001 |
| No. of pharmacies visited to fill L-T4: 2 vs 1 | 4.88 | 4.80 | 4.97 | 4.88 (4.80, 4.97) | <0.0001 |
| No. of pharmacies visited to fill L-T4:>2 vs 1 | 7.15 | 6.97 | 7.34 | 7.15 (6.97, 7.34) | <0.0001 |
| Age group 45–54 vs 18–44 | 1.14 | 1.10 | 1.17 | 1.14 (1.10, 1.17) | <0.0001 |
| Age group 55–64 vs 18–44 | 1.20 | 1.17 | 1.23 | 1.20 (1.17, 1.23) | <0.0001 |
| Age group 65–74 vs 18–44 | 1.26 | 1.23 | 1.30 | 1.26 (1.23, 1.30) | <0.0001 |
| Age group 75+ vs 18–44 | 1.29 | 1.26 | 1.33 | 1.29 (1.26, 1.33) | <0.0001 |
| Female vs Male | 1.02 | 1.00 | 1.04 | 1.02 (1.00, 1.04) | 0.0633 |
| Asian vs White | 1.03 | 0.99 | 1.07 | 1.03 (0.99, 1.07) | 0.1901 |
| Black vs White | 1.03 | 1.00 | 1.06 | 1.03 (1.00, 1.06) | 0.0241 |
| Hispanic vs White | 0.95 | 0.92 | 0.97 | 0.95 (0.92, 0.97) | 0.0001 |
| Unknown race vs White | 1.08 | 1.05 | 1.11 | 1.08 (1.05, 1.11) | <0.0001 |
| Northeast vs Midwest | 0.94 | 0.92 | 0.97 | 0.94 (0.92, 0.97) | <0.0001 |
| South vs Midwest | 0.92 | 0.91 | 0.94 | 0.92 (0.91, 0.94) | <0.0001 |
| West vs Midwest | 0.94 | 0.91 | 0.96 | 0.94 (0.91, 0.96) | <0.0001 |
| Provider Specialty: Other vs Endocrinology | 1.15 | 1.11 | 1.20 | 1.15 (1.11, 1.20) | <0.0001 |
| Provider Specialty: Primary Care vs Endocrinology | 1.09 | 1.06 | 1.13 | 1.09 (1.06, 1.13) | <0.0001 |
| Charlson score | 1.02 | 1.02 | 1.03 | 1.02 (1.02, 1.03) | <0.0001 |
| Atrial fibrillation(Afib) vs Non-Afib | 1.01 | 0.97 | 1.05 | 1.01 (0.97, 1.05) | 0.606 |
| Myocardial infarction(MI) vs Non-MI | 1.07 | 1.00 | 1.14 | 1.07 (1.00, 1.14) | 0.0603 |
| Congestive heart failure(CHF) vs Non-CHF | 1.05 | 1.01 | 1.09 | 1.05 (1.01, 1.09) | 0.0132 |
| Stroke vs Non-stroke | 1.05 | 0.99 | 1.11 | 1.05 (0.99, 1.11) | 0.0942 |
| Thyroid cancer vs Non-thyroid cancer | 1.01 | 0.94 | 1.08 | 1.01 (0.94, 1.08) | 0.8463 |
| Thyroid surgery vs Non-thyroid surgery | 1.22 | 1.13 | 1.31 | 1.22 (1.13, 1.31) | <0.0001 |
| Index year 2009 vs 2008 | 0.93 | 0.87 | 0.99 | 0.93 (0.87, 0.99) | 0.02 |
| Index year 2010 vs 2008 | 0.93 | 0.88 | 0.99 | 0.93 (0.88, 0.99) | 0.0275 |
| Index year 2011 vs 2008 | 1.11 | 1.05 | 1.18 | 1.11 (1.05, 1.18) | 0.0006 |
| Index year 2012 vs 2008 | 1.13 | 1.06 | 1.20 | 1.13 (1.06, 1.20) | <0.0001 |
| Index year 2013 vs 2008 | 1.65 | 1.56 | 1.74 | 1.65 (1.56, 1.74) | <0.0001 |
| Index year 2014 vs 2008 | 2.62 | 2.48 | 2.77 | 2.62 (2.48, 2.77) | <0.0001 |
| Index year 2015 vs 2008 | 1.42 | 1.34 | 1.50 | 1.42 (1.34, 1.50) | <0.0001 |
| Index year 2016 vs 2008 | 1.62 | 1.54 | 1.71 | 1.62 (1.54, 1.71) | <0.0001 |
| Index year 2017 vs 2008 | 2.22 | 2.10 | 2.34 | 2.22 (2.10, 2.34) | <0.0001 |
| Index year 2018 vs 2008 | 3.32 | 3.14 | 3.51 | 3.32 (3.14, 3.51) | <0.0001 |
Table 3.
Diagnostic and procedure codes
| Condition | Codes |
|---|---|
|
| |
| Hyperthyroidism | 242.00, 242.01, 242.10, 242.11, 242.20, 242.21, 242.30, 242.31, 242.4, 242.40, 242.41, 242.8, 242.80, 242.81, 242.9, 242.90, 242.91, 245.9, 245.8, 245.0, 245.1, E05.x |
| Hypopituitarism | 253.7. 253.2, E23.0, E89.3 |
| Pregnacy | V22.x, V23.x, V24.x, V27.x, Z34.x, O09.x, Z3A.x (Exclude Z32.02, V72.41) |
Fig. 2.
Study flow chart
Fig. 3.
ROC curve for logistic regression model of switching from one generic L-thyroxine preparation to another
Footnotes
Compliance with ethical standards
Conflict of interest JSR has received research support through Yale University from Johnson and Johnson, from Medtronic, Inc., the Food and Drug Administration (FDA) (U01FD004585), from the Blue Cross Blue Shield Association, from the Centers of Medicare and Medicaid Services (CMS) (HHSM-500–2013-13018I), from the Agency for Healthcare Research and Quality (R01HS022882), from the National Heart, Lung and Blood Institute of the National Institutes of Health (NIH) (R01HS025164), and from the Laura and John Arnold Foundation. NDS has received research support through Mayo Clinic from the Centers of Medicare and Medicaid Innovation, from the Agency for Healthcare Research and Quality (R01HS025164; R01HS025402; R03HS025517; U19HS024075), from the National Heart, Lung and Blood Institute of the National Institutes of Health (NIH) (R56HL130496; R01HL131535), National Science Foundation, and from the Patient Centered Outcomes Research Institute (PCORI). KJL receives support from the Centers of Medicare and Medicaid Services (CMS) and the National Institute on Aging and the American Federation of Aging Research through the Paul Beeson Career Development Award (K23AG048359). RCS receives support from the Alfred D. and Audrey M. Petersen Professorship in Cancer Research. J.P.B., Y.D., N.H.C., D.J.G., Y.Q., E.R., Z.W., L.Z. have nothing to declare.
Data availability
All data generated or analyzed during this study are included in this published article.
References
- 1.Fuentes AV, Pineda MD, Venkata KCN, Comprehension of top 200 prescribed drugs in the US as a resource for pharmacy teaching, training and practice. Pharmacy (2018). 10.3390/pharmacy6020043 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ross JS, Rohde S, Sangaralingham L, Brito JP, Choi L, Dutcher SK, Graham DJ, Jenkins MR, Lipska KJ, Mendoza M, Qiang Y, Wang Z, Wu Y, Yao X, Shah ND, Generic and brand-name thyroid hormone drug use among commercially insured and medicare beneficiaries, 2007 through 2016. J. Clin. Endocrinol. Metab. (2019). 10.1210/jc.2018-02197 [DOI] [PubMed] [Google Scholar]
- 3.Jonklaas J, Bianco AC, Bauer AJ, Burman KD, Cappola AR, Celi FS, Cooper DS, Kim BW, Peeters RP, Rosenthal MS, Sawka AM, American Thyroid Association Task Force on Thyroid Hormone Replacement. Guidelines for the treatment of hypothyroidism: prepared by the American Thyroid Association Task Force on thyroid hormone replacement. Thyroid (2014). 10.1089/thy.2014.0028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.U.S. Food & Drug Administration, Levothyroxine sodium product information (2019). https://www.fda.gov/drugs/postmarket-drug-safety-information-patients-and-providers/levothyroxine-sodium-product-information. Accessed 21 April 2021
- 5.U.S. Food & Drug Administration, Real-world evidence from a narrow therapeutic index product (Levothyroxine) reflects the therapeutic equivalence of generic drug products (2020). https://www.fda.gov/drugs/news-events-human-drugs/real-world-evidence-narrow-therapeutic-index-product-levothyroxine-reflects-therapeutic-equivalence. Accessed 21 April 2021
- 6.Wallace PJ, Shah ND, Dennen T, Bleicher PA, Crown WH, Optum Labs: building a novel node in the learning health care system. Health Affairs (2014). 10.1377/hlthaff.2014.0038 [DOI] [PubMed] [Google Scholar]
- 7.Optum, Real World Health Care Experiences from Over 150 Million Unique Individuals Since 1993 (2015). https://www.optum.com/content/dam/optum/resources/productSheets/5302_Data_Assets_Chart_Sheet_ISPOR.pdf. Accessed 2 Feb 2021
- 8.Deyo RA, Cherkin DC, Ciol MA, Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. (1992). 10.1016/0895-4356(92)90133-8 [DOI] [PubMed] [Google Scholar]
- 9.Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA, Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med. Care. (2005). 10.1097/01.mlr.0000182534.19832.83 [DOI] [PubMed] [Google Scholar]
- 10.Hennessey JV et al. The Association Between Switching from Synthroid((R)) and clinical outcomes: US evidence from a retrospective database analysis. Adv. Ther. 38(1), 337–349 (2021) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Khandelwal N, Johns B, Hepp Z, Castelli-Haley J, The economic impact of switching from Synthroid for the treatment of hypothyroidism. J. Med. Econ. 21(5), 518–524 (2018). 10.1080/13696998.2018.1443110 [DOI] [PubMed] [Google Scholar]
- 12.Katz M, Scherger J, Conard S, Montejano L, Chang S, Healthcare costs associated with switching from brand to generic levothyroxine. Am. Health Drug Benefits 3(2), 127–134 (2010) [PMC free article] [PubMed] [Google Scholar]
- 13.Endocrine Society, Bioequivalence of Sodium Levothyroxine: An Endocrine Society Position Statement (2008). https://www.endocrine.org/-/media/endosociety/files/advocacyand-statements/all/lt4positionstatementwithmembercommentsheader.pdf?la=en. Accessed 28 Jan 2021
- 14.Thomas K, Kaplan S, Hurricane damage in puerto rico leads to fears of drug shortages nationwide (The New York Times, 2017). https://www.nytimes.com/2017/10/04/health/puerto-rico-hurricane-maria-pharmaceutical-manufacturers.html. Accessed 2 Feb 2021
- 15.Elsevier Clinical Solutions, The impact of rising generic drug prices on the U.S. drug supply chain (2015). http://www.ncpa.co/pdf/elsevier_wp_genericdrug.pdf. Accessed 2 Feb 2021
- 16.Papoian V, Ylli D, Felger EA, Wartofsky L, Rosen JE, Evaluation of thyroid hormone replacement dosing in overweight and obese patients after a thyroidectomy. Thyroid (2019). 10.1089/thy.2019.0251 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data generated or analyzed during this study are included in this published article.