Discrete Choice Research Articles

Public Preference for Off-Label Use of Drugs for Cancer Treatment and Relative Importance of Associated Adverse Events: A Discrete Choice Experiment and Best-Worst Scaling.

Patients may get more treatment options with off-label use of drugs while exposed to unknown risks of adverse events. Little is known about the public or demand-side perspective on off-label drug use, which is important to understand how to use off-label treatment and devise financial assistance. This study aimed to quantify public preference for off-label cancer treatment outcomes, process, and costs, and perceived importance of associated adverse events. A discrete choice experiment and a best-worst scaling were conducted in Hong Kong in December 2022. Quota sampling was used to randomly select the study sample from a territory-wide panel of working-age adults. Preferences and willingness to pay (WTP) for treatment effectiveness, risk of adverse events, mode of drug administration, and availability of off-label treatment guidelines were estimated using a random parameter logit model and latent class model. The relative importance of different adverse events was elicited using Case 1 best-worst scaling. A total of 435 respondents provided valid responses. In the discrete choice experiment, the respondents indicated that extra overall survival as treatment effectiveness (WTP: HK$448,000/US$57,400 for 12-month vs 3-month extra survival) was the most important attribute for off-label drugs, followed by the risk of adverse events (WTP: HK$318,000/US$40,800 for 10% chance to have adverse event vs 55%), mode of drug administration (WTP: HK$42,000/US$5300 for oral intake vs injection), and availability of guidelines (WTP: HK$31,000/US$4000 for available versus not available). Four groups with distinct preferences were identified, including effectiveness oriented, off-label use refusal, oral intake oriented, and adverse event risk aversion. In the best-worse scaling, hypothyroidism, nausea/vomiting, and arthralgia/joint pain were the three most important adverse events based on the perceptions of respondents. Risk-averse respondents, who were identified from the discrete choice experiment, had different perceived importance of the adverse events compared with those with other preferences. Knowing the preference and WTP for cancer treatment-related characteristics from a societal perspective facilitates doctors' communications with patients on decision making and treatment goal-setting for off-label treatment, and enables devising financial assistance for related treatments. This study also provides important insight to inform evaluations of public acceptance and information dissemination in drug development as well as future economic evaluations.

High-dose chemotherapy for patients with stage III breast cancer with homologous recombination deficiency: a discrete choice experiment among healthcare providers.

High-dose chemotherapy with autologous stem cell rescue (HDCT) is currently under investigation as a potential therapy for patients with stage III HER2-negative breast cancer with homologous recombination deficiency (HRD). In addition to survival, the impact on short- and long-term side effects might influence the uptake of HDCT by healthcare professionals. As part of the SUBITO trial, we investigated healthcare professionals' treatment (outcome) preferences for patients with HRD stage III HER2-negative breast cancer and established how healthcare professionals make trade-offs between these treatment outcomes. We conducted a discrete choice experiment in which healthcare professionals were asked to choose repeatedly between scenarios with two treatment options (HDCT or standard of care (SOC)) that varied in outcome with respect to 10-year overall survival (OS), short-term toxicity, long-term cognitive impairment, and premature menopause. We analysed treatment preferences, relative importance, and trade-offs using a multinomial logistic model. Thirty-five of the 151 dedicated breast cancer professionals with extensive experience in treating breast cancer patients completed the survey. The 10-year OS and long-term cognitive impairment were the most important attributes. The results indicate a requirement of 10.4% and 25.1% absolute additional improvement in the 10-year survival rate to justify accepting moderate or severe long-term cognitive impairment as a trade-off, respectively. Therefore, we found in our dataset that healthcare professionals expected a large improvement in 10-year OS to accept moderate to severe cognitive impairment. This information calls for further research into chemotherapy-related cognitive impairment, shared decision-making, and treatment preferences for patients with stage III breast cancer.

Preferences for coordinated care for rare diseases: discrete choice experiment

BackgroundEvidence suggests that coordination of care for people affected by rare diseases is poor. In order to improve the way that care is coordinated it is necessary to understand the preferences of people affected by these conditions, and providers. The aim of this study was to examine patient, parent and carer, and health care professional preferences for different attributes of care coordination for people affected by rare diseases. We conducted a discrete choice experiment using online surveys. There were no restrictions on participants in terms of rare conditions, demographic factors other than age, or geographical location within the UK. Choice scenarios were based on the following attributes: annual cost of attending appointments; access to health records; access to clinical expertise; support of a care coordinator; access to a specialist centre; and, the existence of a documented plan for emergency care. Data were analysed using alternative-specific conditional logit regression models.ResultsValid responses were obtained from 996 individuals (528 patients, 280 carers, 188 health care professionals) between August and December 2019. All attributes significantly influenced the type of service respondents preferred. Patients, carers and health professionals’ preferences for care coordination were influenced by: the cost of attending appointments; access to health records; clinical expertise; role of care coordinators; access to specialist centres; and the existence of plan for emergency care. There were no statistically significant differences in the preferences between patients and carers. Preferences of health professionals differed to those of patients and carers. Both patients and carers selected responses which granted them a greater degree of autonomy in relation to the role of care coordinators, whereas health professionals preferred services where care coordinators had more autonomy. Health care professionals also expressed a stronger preference for a documented formal emergency plan to be in place.ConclusionsThe findings highlight that people value better coordinated care, in line with policy documents emphasising commitments to coordinated care for people affected by rare diseases. This study highlights the factors that could be included in service provision as ways of improving the coordination of care for people affected by rare diseases.

La scelta dell’ausiliare nei tempi composti: Variazione nell'interlingua di apprendenti ungheresi e inglesi di italiano L2/LS

The aim of this study is to identify some factors that determine variability in the interlanguage of L2/FL learners of Italian concerning auxiliary selection (essere ‘to be’ vs avere ‘to have’) in compound past tenses. Data were collected by means of a grammaticality judgement test and a binary choice sentence completion test. The participants were L1 English and Hungarian university students of various levels of proficiency (A2, B1, B2 and C1). The results were interpreted both in correlation with certain learning variables such as L1, proficiency level, and the category the main verbs belong to according to Sorace’s Auxiliary Selection Hierarchy. Based on the results, some hypotheses are put forward on teaching material design and on the possibility of harmonising the didactic intervention and the acquisitional sequence observed.

From policy to practice: The role of national policy instruments and social barriers in UK energy efficiency adoption in households

This paper investigates the adoption trends for energy efficiency (EE) measures in the UK residential sector and the extent to which national-level EE policies impact households' willingness to adopt EE measures and technologies. Using survey data from a representative sample of more than 2000 households across the UK and discrete choice models, this research examines the preferences given to different policy schemes such as tax rebates, grants, loans, easing administrative barriers or information campaigns; and its role on household decisions to adopt EE measures. We find that national-level policies in the form of grants and subsidies, information campaigns on savings and a reduction of administrative barriers would positively impact the probability of a household to adopt such measures. Differences are found between the policy drivers in the adoption of low-cost EE measures e.g. loft insulation vs. High-cost EE measures e.g. heat pumps. Importantly, our results highlight the role of policy mixes in fostering the adoption of EE measures in UK. In combination with other policies, tax credits can be effective in fostering the adoption on EE measures. Our findings have implications for policymakers seeking to enhance the effectiveness of EE policies and address social challenges to promote sustainable energy use.

Spatiotemporal dynamics of locomotor decisions in Drosophila melanogaster.

Decision-making in animals often involves choosing actions while navigating the environment, a process markedly different from static decision paradigms commonly studied in laboratory settings. Even in decision-making assays in which animals can freely locomote, decision outcomes are often interpreted as happening at single points in space and single moments in time, a simplification that potentially glosses over important spatiotemporal dynamics. We investigated locomotor decision-making in Drosophila melanogaster in Y-shaped mazes, measuring the extent to which their future choices could be predicted through space and time. We demonstrate that turn-decisions can be reliably predicted from flies' locomotor dynamics, with distinct predictability phases emerging as flies progress through maze regions. We show that these predictability dynamics are not merely the result of maze geometry or wall-following tendencies, but instead reflect the capacity of flies to move in ways that depend on sustained locomotor signatures, suggesting an active, working memory-like process. Additionally, we demonstrate that fly mutants known to have sensory and information-processing deficits exhibit altered spatial predictability patterns, highlighting the role of visual, mechanosensory, and dopaminergic signaling in locomotor decision-making. Finally, highlighting the broad applicability of our analyses, we generalize our findings to other species and tasks. We show that human participants in a virtual Y-maze exhibited similar decision predictability dynamics as flies. This study advances our understanding of decision-making processes, emphasizing the importance of spatial and temporal dynamics of locomotor behavior in the lead-up to discrete choice outcomes.

Movement decisions driving metapopulation connectivity respond to social resources in a long-lived ungulate, bighorn sheep (Ovis canadensis).

The spatial availability of social resources is speculated to structure animal movement decisions, but the effects of social resources on animal movements are difficult to identify because social resources are rarely measured. Here, we assessed whether varying availability of a key social resource-access to receptive mates-produces predictable changes in movement decisions among bighorn sheep in Nevada, the United States. We compared the probability that males made long-distance 'foray' movements, a critical driver of connectivity, across three ecoregions with varying temporal duration of a socially mediated factor, breeding season. We used a hidden Markov model to identify foray events and then quantified the effects of social covariates on the probability of foray using a discrete choice model. We found that males engaged in forays at higher rates when the breeding season was short, suggesting that males were most responsive to the social resource when its existence was short lived. During the breeding season, males altered their response to social covariates, relative to the non-breeding season, though patterns varied, and age was associated with increased foray probability. Our results suggest that animals respond to the temporal availability of social resources when making the long-distance movements that drive connectivity. This article is part of the theme issue 'The spatial-social interface: a theoretical and empirical integration'.

Cannabis consumers’ preferences for legal and illegal cannabis: evidence from a discrete choice experiment

BackgroundIn U.S. states that legalized and commercialized recreational cannabis, cannabis sales in illegal markets are still sizable or even larger than those in legal markets. This study aimed to assess cannabis consumers’ preferences for purchasing cannabis from legal and illegal markets and estimate the trade-offs under various policy scenarios.Methods963 adults were recruited, who used cannabis in the past year and lived in a state with recreational cannabis legalization. In a discrete choice experiment, participants chose purchasing cannabis from a legal dispensary or an illegal dealer with varying levels in product attributes including quality, safety, accessibility, potency, and price. Mixed logit models were used to analyze preferences.ResultsThe likelihood of choosing legal cannabis increased with a higher quality, the presence of lab test, a shorter distance to seller, a higher tetrahydrocannabinol level, and a lower price. The likelihood of choosing illegal cannabis increased with a higher quality, a shorter distance to seller, and a lower price. Among product attributes, quality and accessibility were perceived to be the most important for legal cannabis and price was perceived to be the most important for illegal cannabis. Policy simulations predicted that improving quality, ensuring safety, allowing delivery services, increasing dispensary density, and lowering prices/taxes of legal cannabis may reduce illegal cannabis market share.ConclusionsIn the U.S., cannabis consumers’ preferences for illegal cannabis were associated with both legal and illegal cannabis product attributes. Policies regulating legal cannabis markets should consider potential spillover effects to illegal markets.

Valuation of the EQ-5D-3L in Jordan.

In Jordan, no national value set is available for any preference-accompanied health utility measure. This study aims to develop a value set for EQ-5D-3L based on the preferences of the Jordanian general population. A representative sample of the Jordanian general population was obtained through quota sampling involving age, gender, and region. Participants aged above 18 years were interviewed via videoconferencing using the EuroQol Valuation Technology 2.1 protocol. Participants completed ten composite time trade-offs (cTTO) and ten discrete choice experiments (DCE) tasks. cTTO and DCE data were analyzed using linear and logistic regression models, respectively, and hybrid models were applied to the combined DCE and cTTO data. A total of 301 participants with complete data were included in the analysis. The sample was representative of the general population regarding region, age, and gender. All model types applied, that is, random intercept model, random intercept Tobit, linear model with correction for heteroskedasticity, Tobit with correction for heteroskedasticity, and all hybrid models, were statistically significant. They showed logical consistency in terms of higher utility decrements with more severe levels. The hybrid model corrected for heteroskedasticity was selected to construct the Jordanian EQ-5D-3L value set as it showed the best fit and lowest mean absolute error. The predicted value for the most severe health state (33333) was -0.563. Utility decrements due to mobility had the largest weight, followed by anxiety/depression, while usual activities had the smallest weight. This study provides the first EQ-5D-3L value set in the Middle East. The Jordanian EQ-5D-3L value set can now be used in health technology assessments for health policy planning by the Jordanian health sector's decision-makers.

A Reporting Checklist for Discrete Choice Experiments in Health: The DIRECT Checklist.

Reporting standards of discrete choice experiments (DCEs) in health have not kept pace with the growth of this method, with multiple reviews calling for better reporting to improve transparency, assessment of validity and translation. A key missing piece has been the absence of a reporting checklist that details minimum standards of what should be reported, as exists for many other methods used in health economics. This paper reports the development of a reporting checklist for DCEs in health, which involved a scoping review to identify potential items and a Delphi consensus study among 45 DCE experts internationally to select items and guide the wording and structure of the checklist. The Delphi study included a best-worst scaling study for prioritisation. The final checklist is presented along with guidance on how to apply it. This checklist can be used by authors to ensure that sufficient detail of a DCE's methods are reported, providing reviewers and readers with the information they need to assess the quality of the study for themselves. Embedding this reporting checklist into standard practice for health DCEs offers an opportunity to improve consistency of reporting standards, thereby enabling transparency of review and facilitating comparison of studies and their translation into policy and practice.

Understanding author choices in the current conservation publishing landscape.

Conservation literature addresses a broad spectrum of interdisciplinary questions and benefits. Conservation science benefits most when a diverse range of authors are represented, particularly those from countries where much conservation work is focused. In other disciplines, it is well known that barriers and biases exist in the academic publishing sphere, which can affect research dissemination and an author's career development. We used a discrete choice experiment to determine how 7 journal attributes affect authors' choices of where to publish in conservation. We targeted authors directly by contacting authors published in 18 target journals and indirectly via communication channels for conservation organizations. We only included respondents who had previously published in a conservation-related journal. We used a multinomial logit model and a latent class model to investigate preferences for all respondents and distinct subpopulations. We identified 3 demographic groups across 1038 respondents (older authors from predominantly middle-income countries, younger authors from predominantly middle-income countries, and younger authors from high-income countries) who had published in conservation journals. Each group exhibited different publishing preferences. Only 2 attributes showed a consistent response across groups: cost to publish negatively affected journal choice, including authors in high-income countries, and authors had a consistent preference for double-blind review. Authors from middle-income countries were willing to pay more for society-owned journals, unlike authors from high-income countries. Journals with a broad geographical scope that were open access and that had relatively high impact factors were preferred by 2 of the 3 demographic groups. However, journal scope and open access were more important in dictating journal choice than impact factor. Overall, different demographics had different preferences for journals and were limited in their selection based on attributes such as open access policy. However, the scarcity of respondents from low-income countries (2% of respondents) highlights the pervasive barriers to representation in conservation research. We recommend journals offer double-blind review, reduce or remove open access fees, investigate options for free editorial support, and better acknowledge the value of local-scale single-species studies. Academic societies in particular must reflect on how their journals support conservation and conservation professionals.

Weak identification in discrete choice models

We study the impact of weak identification in discrete choice models, and provide insights into the determinants of identification strength in these models. Using these insights, we propose a novel test that can consistently detect weak identification in commonly applied discrete choice models, such as probit, logit, and many of their extensions. Furthermore, we demonstrate that when the null hypothesis of weak identification is rejected, Wald-based inference can be carried out using standard formulas and critical values. A Monte Carlo study compares our proposed testing approach against commonly applied weak identification tests. The results simultaneously demonstrate the good performance of our approach and the fundamental failure of using conventional weak identification tests for linear models in the discrete choice model context. Lastly, we apply our approach in two empirical examples: married women labor force participation, and US food aid and civil conflicts.

A large-scale investigation of eating behaviors and meal perceptions in Italian primary school cafeterias: The relationship between emotions, meal perception, and food waste

We explored the relationships between meal perception, emotions, and food waste in different age groups in primary school cafeterias. The aims were to assess the effect of age and gender on perceptions of the cafeteria’s meals, emotions, and food waste, and to assess the relationships among food waste, meal perception, and emotions. 1,092 children (5–11 years old [y.o.], 53 % females) completed a questionnaire in nine Italian schools. The main collected variables were the pleasure of eating in the cafeteria, declared liking for different foods, emotional responses to eating in the cafeteria, meal description evaluated as binary choices considering different aspects (duration, presence of noise, perceived comfort, crowding, fun, possibility of movement), and food waste (overall and for different foods). Age but not gender significantly affected meal perception, emotions, and food waste. The general attitude towards eating in the cafeteria was more positive among the youngest children (6–7 y.o.), while it steadily worsened from 8 y.o. on. Pasta, meat, and fruits were consistently the most liked foods, fish, vegetables, and legumes the least. Mealtime in the cafeteria was generally described as noisy and crowded, but also comfortable and fun. Children used negative emojis less to describe the mealtime. Food waste was high: 22 % of children declared leaving food on the plate “frequently” and 8 % “always.” However, lower food waste was associated with a positive emotional status and a positive meal perception. Intervention actions should improve the emotional response and the lunch environmental context in school cafeterias.

Impact of bedaquiline resistance probability on treatment decision for rifampicin-resistant TB

<sec><title>BACKGROUND</title>Accurate diagnosis of bedaquiline (BDQ) resistance remains challenging. A Bayesian approach expresses this uncertainty as a probability of BDQ resistance (prBDQR) with a 95% credible interval. We investigated how prBDQR information influences BDQ prescribing decisions.</sec><sec><title>METHOD</title>We performed a discrete choice experiment with 55 international rifampicin-resistant tuberculosis physicians. We employed mixed-effects multinomial logistic regression to quantify the effect of prBDQR, patient attributes, and contextual factors on the decision to continue BDQ or not when sequencing results become available.</sec><sec><title>RESULTS</title>PrBDQR was the most influential factor for BDQ decision-making, three times greater than treatment response. Each percentage point increase in prBDQR resulted in 8.2% lower odds (OR 0.92, 95% CI 0.90–0.93) of continuing BDQ as a fully effective drug and 5.0% lower odds (OR 0.95, 95% CI 0.94–0.96) of continuing it but not counting it as an effective drug. The most favourable patient profile for prescribing BDQ as a fully effective drug was a patient receiving the BPaLM regimen (BDQ, pretomanid, linezolid and moxifloxacin) with low prBDQR, good 1-month treatment response, fluoroquinolone-susceptible TB, and no prior BDQ treatment. Physicians with higher discomfort with uncertainty and more years of experience with BDQ were more inclined to stop BDQ.</sec><sec><title>CONCLUSION</title>Given the uncertainty of genotype-phenotype associations, physicians valued prBDQR for BDQ decision-making in rifampicin-resistant TB treatment.</sec>

Assessing the substitutability of mobile and fixed internet: The impact of 5G services on consumer valuation and price elasticity

In this study, we explore the dynamics of consumer choices in the Polish telecommunications market, focusing on preferences and valuations for home fixed, home mobile, and purely mobile Internet connections. Key attributes such as speed, latency, data limits, and cost are examined. Central to our research is the investigation of how the integration of 5G technology might influence demand elasticity. Using a detailed discrete choice experiment, we apply a mixed logit model with random parameters to analyze stated choice data, enabling us to unravel the complexities of demand elasticity, especially in terms of own and cross-price elasticities. This approach facilitates an assessment of the degree of substitutability between fixed and mobile Internet services.Our findings indicate a moderate substitution effect between fixed and mobile Internet services. Results from a Small but Significant and Non-transitory Increase in Price (SSNIP) test suggest that these markets should continue to be regulated separately, mirroring the distinct regulation observed in fixed and mobile telephony. Furthermore, simulations provide insights into potential future market shifts with the advent of 5G services. This paper contributes significantly to the discourse on fixed-mobile Internet substitution and offers vital insights for defining markets in antitrust discussions, competitive agreements, and potential mergers within the telecom sector.

Exploring primary care physician biases in adolescent contraceptive counseling

Objective(s)While previous literature has shown clinician bias in adult contraceptive counseling, less is known on the biases clinicians may exhibit when counseling adolescents about contraception. Our study aimed to describe long-acting reversible contraception (LARC) counseling and prescribing practices of adolescent-serving clinicians. Study designThis study used a cross-sectional discrete choice experiment mixed methods design. We sent a survey containing vignettes and items pertaining to demographics and beliefs to a convenience sample of adolescent-serving clinicians across the United States. ResultsOf 296 clinicians, 80% were in pediatrics, and had geographic, practice setting, gender, and racial diversity. Most clinicians reported being up-to-date with current literature regarding contraception. Sixty-eight percent of respondent’s practices administer contraceptive injections, but only 17% place intrauterine devices (IUDs). Of those who do insert IUDs, nearly half inserted five or fewer within the last year. Patients’ younger age and Hispanic ethnicity were associated with lower odds, and history of pregnancy was associated with higher odds, of clinicians’ recommending LARC. Across all vignettes, the top five reasons clinicians chose their first-choice method for the patient were adherence or compliance, efficacy, side effects, patient age, and reversibility. Conclusion(s)Clinicians often recommend contraceptives based on adherence, efficacy, and age, and we found younger age, minoritized race or ethnicity, and history of pregnancy were all associated with LARC recommendations, indicating potential biases against teen parents and assumptions about adolescents’ priorities. This may impede the provision of patient-centered contraceptive counseling for adolescents, and highlights the need for improved education and practice changes post-Dobbs. ImplicationsWe found that clinicians demonstrated several biases in how they provide contraceptive recommendations to adolescent patients. These biases were often associated with their personal beliefs and experiences. Our findings can guide the development of future interventions aimed at improving adolescent reproductive health counseling and care delivery in primary care settings.

Choosing Your Sports Partners: Assessing Selection Preferences Through Observational and Experimental Studies

ABSTRACT Previous research has demonstrated the positive impact of doing sports together on maintaining physical activity levels. Yet, there remains a gap in understanding the selection processes leading to sports partnerships. We address this question using a two-study design. Study 1 examines longitudinal ego-network data; we analyze data of 1,222 sports partnerships among 409 Dutch university students. We find that emotional closeness, rather than skill similarity, is a key determinant of sports partnership continuation over time. We do not find that the sports setting in which sports partnerships are embedded (i.e. sports clubs, informal groups, commercial gyms, unorganized settings) affect the stability of sports partnerships. Study 2 employs a discrete choice experiment, in which individuals are tasked to choose a sports partner out of choice-sets consisting of three sports partners who vary in the types of motivations and resources they offer (i.e. social comparison, knowledge, companionship, encouragement). People express a strong preference for sports partners who emphasize social aspects of sports, but as individuals become more active, sports partners who emphasize social comparison and who are knowledgeable about sports gain attractiveness. We discuss the implications of these findings for tailored strategies leveraging social networks to promote sports participation.

V-112.3: Preferences for the treatment of polyomavirus (BKPyV) infection in kidney transplant recipients: A discrete choice experiment with patients, caregivers and clinicians.

V-112.3: Preferences for the treatment of polyomavirus (BKPyV) infection in kidney transplant recipients: A discrete choice experiment with patients, caregivers and clinicians.

Assessing the Deterrent Effects of Ignition Interlock Devices

IntroductionIgnition interlock devices installed after conviction for driving under the influence of alcohol (DUI) have been shown to reduce subsequent DUI arrests (specific deterrence). However, there is little evidence on how interlock-device penalties might affect general deterrence, that is, deterring people from driving after consuming alcohol prior to a DUI conviction. MethodsA discrete choice experiment was conducted and data were analyzed in 2023 with 583 U.S.-based adults who consume alcohol at least once in the past week to assess the deterrent effects of five different penalties (fine, jail time, interlock device, license suspension, alcohol treatment) for alcohol-impaired driving under randomized sequential scenarios of high (20% chance of being caught) and low (1%) police enforcement. Participants resided in 46 states. ResultsDeterrent effects of an interlock penalty, operationalized as having to install an interlock device for one year, are large and on par with a twentyfold increase in police enforcement activity (from 1% chance of being caught to 20%), or a $2,000 increase in the DUI fine under the status quo enforcement regime. On average, a 1-year interlock penalty had the same deterrent effect as a 10-day increase in jail time. ConclusionsWider use of interlock devices as a DUI penalty could have large deterrent effects, independent of their ability to physically prevent the motor vehicle of an intoxicated driver from starting. The deterrent effect documented here adds to evidence on interlock devices’ overall effectiveness as well as their potential to shift DUI penalties away from criminalization (jail time) and towards immobilization and rehabilitation.

Evaluating the Preferences and Willingness-to-Pay for Oral Antidiabetic Drugs Among Patients with Type 2 Diabetes Mellitus in China: A Discrete Choice Experiment.

To quantify the preferences for an oral antidiabetic drug (OAD) among patients with type 2 diabetes mellitus (T2DM) in China. A discrete choice experiment (DCE) with hypothetical OAD profiles was performed among patients with T2DM recruited from both online and offline sources. Each patient completed 12 DCE choice tasks. The attributes, elicited through mixed methods, include blood glucose level decrease, blood glucose level stability, frequency of medication, gastrointestinal side effects, dose adjustment and out-of-pocket expense. The conditional logit regression model was used to analyze the data. Patients' willingness-to-pay (WTP) was also calculated. Subgroup analyses based on patient characteristics were also conducted. A total of 741 respondents were included in the analysis sample, covering 456 respondents online and 285 offline. The result showed that all attributes and levels were statistically significant, except one level "dose adjustment required for patients with hepatic or renal insufficiency" in the attribute of dose adjustment. WTP results showed that patients were willing to pay 12.06 and 23.20 yuan, respectively to reduce the frequency of medication from "once per day" and "three times per day" to "once every 2 weeks", respectively. Subgroup analyses showed that the frequency of medication (once versus two to three times per day) had the largest impact and influenced most coefficient estimates. The results suggest that Chinese patients with T2DM prioritized better efficacy, less frequency of medication, lower gastrointestinal side effects, no dose adjustment required for patients with hepatic or renal insufficiency, and less out-of-pocket expense of OAD treatment.