This article provides the first controlled evaluation of how different information materials explaining the risks from radon influenced people's perceptions of these risks. Using a panel study, it was possible to observe how stated risk perceptions responded to information about indoor radon concentrations and brochures explaining the radon readings. The findings indicate that risk communication policies can be effective in modifying risk perceptions. Moreover, they have three specific implications for radon policy: (1) Public officials should not adopt strategies that provide minimal risk information to the public as a means of avoiding undue alarm, for this can have the reverse effect; (2) measures of the effectiveness of risk communication will depend on how education and behavior change are defined; (3} categorical guidelines about risk without quantitative information can lead people to treat the levels as thresholds, creating an artificial discontinuity in their responses to small changes in risk perceptions.

To quantify value of hope as a construct for potential use in value assessment frameworks. We designed a discrete-choice experiment in which pairs of treatment alternatives were constructed using probabilities of achieving long-term (i.e. 10-year) survival (0%, 5%,10%, and 20%) and the corresponding probabilities and durations of short-term survival such that expected survival for the alternative equaled 2, 3, or 5 years. Other attributes were health status and out-of-pocket cost, representing 0%, 10%, 30% or 80% of participants’ reported annual income. Each participant was randomly assigned to one of 12 blocks of 10 choice questions. Following a pilot study in 120 adults, 200 patients with cancer or history of cancer were recruited by Cancer Support Community. We computed dollar-equivalent values for treatments offering 5%, 10% and 20% chances of long-term survival versus a 0% chance of long-term survival (i.e. guaranteed 2-, 3- or 5-year survival). We evaluated preference heterogeneity using mixed-logit and latent-class models. Participants’ choices implied that they preferred a 5% versus a 0% chance of long-term survival (p=0.054) and a 10% versus a 0% chance of long-term survival (p=0.001) when holding expected survival constant. However, participants did not indicate a preference for treatments offering a 20% versus a 0% chance of long-term survival (p=0.38) due to the offsetting 80% chance of shorter survival to maintain the same duration of expected survival. Compared to a 0% chance, dollar-equivalent values were $5,975 (95%CI: -145 to 12,629) for a 5% chance and $12,421 (95%CI: 5,082 to 20,744) for a 10% chance of long-term survival, independent of health status or expected survival. Latent-class analysis revealed 4 groups with distinct preference patterns. Independent of expected survival and health status, patient preferences are consistent with positive value for hope. However, this finding does not universally hold in all situations nor across individuals.

vant stakeholders on an international level, a list of provisional outcomes for CMN was critically reviewed through an online questionnaire and further modified to provide clear definitions for every outcome. Thereafter, various stakeholders discussed the inclusion of potential outcomes in an online consensus conference. Finally, ten outcomes were selected for the COS in two online votings, nine of which applied to both care and research: (i) skin appearance: ‘size of CMN’, ‘colour of CMN’, ‘texture of CMN’ and ‘satellite naevi number’; (ii) quality of life: ‘emotional distress’; (iii) neoplasm: ‘presence of melanoma’; (iv) neurology: ‘neurological symptoms and signs’; (v) adverse events: ‘wound problems of the CMN’ and ‘scar problems’; and, for research only, (vi) pathology: ‘molecular characteristics’. Nine of ten outcomes were redundant for both patient care and research. This raises the question of whether two distinct COS are needed for this condition. In contrast, outcomes for other clinically relevant domains such as neuroimaging to rule out neurocutaneous melanocytosis and psychological functioning were not included in the final set of outcomes. The past two steps of the OCOMEN project are undoubtedly important milestones in the journey towards a uniform approach to CMN as a condition with high variability in its clinical features and a paucity of data. However, the most difficult step is still looming ahead: to define a core outcome measurement set based on specific measurement instruments. This endeavour is not entirely novel as several classification schemes have been proposed for CMN in the past. Within the OCOMEN initiative, Fledderus et al. agreed on size of the lesion, number of satellite naevi, colour and texture as outcomes for the domain ‘skin appearance’. These parameters were also part of a consensus-based scheme for the categorization of CMN published previously by an interdisciplinary panel of experts. In the next step, we will see how the COS of the OCOMEN initiative will differ from existing recommendations, and what additional value it will generate for physicians and patients. In particular, measuring the domains and outcomes relating to extracutaneous findings such as emotional distress or neurological symptoms and signs can add significant novelty and fill existing gaps as they are not sufficiently considered in current categorization schemes.

New allergy desensitization treatments being developed offer the potential for reductions in risk of severe allergic reactions to accidental peanut exposures. This study aimed to quantify treatment features and decision-maker characteristics that influence stated preferences for desensitization treatment among caregivers of peanut-allergic children in the United States. A national sample of 500 parents completed an online discrete-choice-experiment survey. Respondents evaluated a series of two hypothetical treatment profiles alongside a no-treatment option. Treatment profiles varied in mode of administration, peanut tolerance level, proportion discontinuing treatment due to side effects, and cost. A probit model estimated probability of choosing treatment at least once out of 12 questions, as compared to never choosing treatment. A latent-class model estimated relative-importance weights for treatment features. Approximately 90% chose treatment at least once. Fathers, non-Asians, and younger parents were more likely to choose treatment at least once. Meanwhile, those who reported 100% chance of being able to manage an allergic reaction, those who perceived capsule as ‘very inconvenient’, and those who reported cost as being most important were more likely never to choose treatment. Among the 90% who chose treatment at least once, latent-class analysis identified 3 classes, representing 41%, 33%, and 27% of the sample, respectively. If offered a treatment with the mean value for efficacy, side effects, and cost, the model predicted that 8%, 70%, and 4% would choose treatment for each class, respectively. If offered the best efficacy, no side effects, and lowest cost, the model predicted that 2/3, all, and about half would choose treatment, respectively. Factors affecting likelihood to select peanut-allergy desensitization treatment were heterogeneous among caregivers participating in a choice-experiment study. Understanding the factors affecting likelihood to start desensitization treatment will be important for shared decision making between clinicians and caregivers.

More women with ovarian cancer in remission are receiving maintenance therapy with PARP inhibitors. But, it is unclear whether women are willing to accept side effects and risks associated with maintenance therapy for extended progression-free survival (PFS) or overall survival (OS). We designed a discrete-choice experiment to elicit the preferences of women with ovarian cancer regarding tradeoffs among 6 attributes relevant to the decision for maintenance PARP inhibitor therapy versus surveillance: (1) OS (36, 38, 42 months); (2) PFS (15, 17, 21 months); (3) nausea (none, mild, moderate); (4) fatigue (none, mild, moderate); (5) probability of death from myelodysplastic syndrome/acute myelogenous leukemia (MDS/AML) (0%, 1%, 5%, 10%); and (6) monthly out-of-pocket cost ($0, $50, $500, $1,000). Choice questions represented 2 maintenance scenarios and an opt-out scenario representing a treatment break. We analyzed choice data using random-parameters logit regression. Of 150 women with ovarian cancer recruited, 95 were eligible and completed the survey. Mean age was 62, 48% had recurrent ovarian cancer, and 13% were currently taking a PARP inhibitor. Preference weights were logically ordered, with better clinical outcomes significantly preferred to worse (except between 0% and 1% risk of MDS/AML) and lower costs preferred to higher. On average, participants would tolerate a 2% (95%CI: -1, 5) additional risk of MDS/AML in exchange for 2 additional months of PFS and 7% (4, 10) additional risk of MDS/AML in exchange for 6 additional months of PFS. Risk tolerance was greater for gains in OS, accepting a 6% (2, 8) additional risk of MDS/AML for a 2-month gain in OS and a 13% (10, 20) additional risk for a 6-month OS gain. To realize clinically observed levels of improvement in PFS with maintenance PARPi therapy, women with ovarian cancer are willing to accept a risk of MDS/AML that surpasses clinically observed risk levels.

To quantify benefit-risk tradeoffs relevant to transcatheter mitral-valve repair versus medical therapy for patients with heart failure and symptomatic secondary mitral regurgitation, we conducted a discrete-choice experiment (DCE) survey. The DCE was designed to quantify patients’ tolerance for 30-day mortality or serious bleeding risks to achieve improvements in physical functioning or reductions in heart-failure hospitalizations. In addition to two experimentally designed medical-device options, a fixed opt-out (i.e. no device) option represented an individual with New York Heart Association (NYHA) class IV heart failure symptoms and an expectation of 5 heart-failure hospitalizations in the next 2 years. Participants with heart failure were recruited from the Duke University Health System (n=175) or an online US-based panel (n=244). Across both samples, 56.3% were male, mean age was 65 years, and 44% had symptoms consistent with NYHA class II, and 26.4% with NYHA class III or IV. Nearly one-quarter (23.5%) always chose device profiles offering the higher level of physical functioning despite mortality and bleeding risks as high as 10%. Among respondents who at least once chose a device profile offering a lower level of functioning, physical functioning improvements equivalent to a change from NYHA class IV to III was approximately six times more important than a change from NYHA class III to II. DCE utility gains and losses implied that participants would accept up to a 9.7 percentage-point (95% CI: 8.2%- 13.3%) increase in risk of 30-day mortality with devices offering improved functioning from NYHA class IV to III, but just 2.0% (95% CI: 1.4% to 2.7%) for an improvement from NYHA class III to II. As severity of heart-failure symptoms worsen, patients will accept greater risks to achieve improvements in physical functioning. These findings can inform interpretation of clinical trial findings, clinical decision making and regulatory decisions.

Introduction Quantitative data on patients’ willingness to accept benefit-risk tradeoffs with new medical procedures can provide useful information to regulators and clinicians. We adhered to 2016 Guidance from the FDA's Center for Drugs and Radiological Health to generate evidence on patient preferences relevant to potential risks and benefits with transcatheter mitral-valve repair versus medical therapy for patients with heart failure (HF) and secondary mitral regurgitation. Hypothesis We hypothesized that tolerance for procedure-associated risks would vary across different levels of improvement in physical functioning. Methods We designed a discrete-choice experiment to quantify patients’ tolerance for procedure-associated increases in 30-day mortality or serious bleeding risks to achieve improvements in physical functioning or reductions in HF hospitalizations. After describing the procedure, risks and outcomes, respondents were asked to choose their preferred option among two experimentally-designed procedure profiles or a no-procedure profile for each of 8 choice questions. The survey was administered to two samples: HF patients treated at the Duke University Health System (DUHS) and an online panel of individuals reporting a HF diagnosis. We applied random-effects logit regression to model choices as a function of benefit and risk levels. Results Respondents from DUHS (n=175) were slightly older (mean age 67 vs. 64 yrs, p=0.02) and more likely to be male (62% vs. 52%, p=0.06) than panel respondents (n=244). Across both samples, about 44% had symptoms consistent with NYHA class II, and 26.4% had symptoms consistent with NYHA class III or IV. Both samples revealed similar benefit-risk preferences. Approximately one-quarter (23.5%) chose procedure profiles offering a higher level of physical functioning across all 8 choice questions despite mortality and bleeding risks up to 7% or 10%. Amongst respondents who at least once chose a procedure profile offering a lower level of functioning, an improvement equivalent to moving from NYHA class IV to III was approximately six times more preferred than an improvement from NYHA class III to II. Estimated preference weights revealed that respondents would accept up to a 9.7 percentage-point (95% CI: 8.2%-13.3%) increase in risk of 30-day mortality with procedures that could improve functioning from NYHA class IV to III, but only up to 2.0% (95% CI: 1.4%-2.7%) for an improvement from NYHA class III to II. When controlling for improvements in functioning, the number of HF hospitalizations did not significantly impact respondents’ choices. Conclusions When faced with more severe versus less severe HF symptoms, patients are willing to tolerate greater risks to achieve improvements in physical functioning. These findings can be valuable to regulators. Nevertheless, variability in individual patient preferences underscores the importance of individualized shared decision-making.

Approximately 25% of breast cancer diagnoses detected with mammography represent ductal carcinoma in situ (DCIS), a non-invasive condition that is often aggressively managed with multimodal therapy. Management options may lower risk of invasive disease, but women encounter changes in breast appearance, treatment-associated risks and side effects. We aimed to quantify acceptable tradeoffs among features of alternative outcomes following various management options for DCIS. We designed and fielded a discrete-choice experiment survey among women presenting for screening mammography. The survey required women to complete 10 DCE choice questions representing 2 alternative health states defined by 5 attributes, each with 3 or 4 levels: breast appearance, severity of infection within the first year, years of chronic pain, years of hot flashes and risk of developing/dying from breast cancer within 10 years. Random-parameters logit models were used to fit the relationship between response options and attribute levels. Nearly one-third (31%) of the 150 participants dominated on the risk of invasive disease, always selecting the option with the lowest cancer risk. Among the remaining women, on average, they would accept a 2.9% (95% CI: 1.9- 3.8) absolute 10-year risk of breast cancer to have a normal breast appearance compared to breast appearance after a mastectomy. Also, women would accept a 2.7% (95% CI: 1.9-3.6) risk of breast cancer to have a breast appearance following lumpectomy compared to mastectomy. There was variability in acceptable tradeoffs depending on surgical outcome, with tolerable breast cancer risk levels ranging from 2.2% to 3.5%. The majority of participants (69%) found some increases in breast cancer risk acceptable to minimize changes in breast appearance following surgery. Clinicians treating women for DCIS should recognize that considerable heterogeneity exists in how women prioritize invasive cancer risk relative to breast appearance.

Women with epithelial ovarian cancer (EOC) are appropriate candidates for genetic testing and counseling. However, effective provision and use of test information require understanding the preferences and values of women with EOC. This study aimed to quantify values of tests to identify gene variants with known and unknown significance to women with EOC. Adult EOC women with a referral for genetic testing were recruited to participate in a discrete-choice-experiment survey to elicit their preferences for 5 features of genetic testing: probability of identifying a deleterious mutation (60%, 80%, or 88%), probability of identifying a variant of uncertain significance (VUS) (5%, 20%, or 40%), sample requirements (blood or saliva), turn-around time (1, 2 or 4 weeks), and out-of-pocket cost ($0, $100, $250, or $1,000). Subjects viewed an educational video describing the 5 test features followed by a series of 10 choices between pairs of constructed genetic tests with varying levels of each feature. Random-parameters logit was used to estimate relative important weights for all feature levels. Money-equivalent values of various features were calculated in secondary analysis. On average, the 93 EOC women who completed the survey logically placed the greatest importance on the ability of tests to detect a deleterious mutation. However, VUS detection was next most important. Over the ranges included in the study, detection of a deleterious mutation was twice as important as detection of a VUS. For example, women valued an improvement in detection probability from 60% to 88% in a deleterious mutation at $877 (CI: $583-$1226) and an improvement in detecting a VUS from 5% to 40% at $492 (CI: $305-$718). Women placed considerable value on incidental, non-actionable VUS detection. The expectation that future research could make VUS results valuable suggests it may be an important topic for genetic counseling.

To investigate whether there are groups of older adults who express different preferences for treatment benefits and risks for delaying the onset of Alzheimer’s disease (AD). A total of 1004 older adults in the US completed a web-enabled discrete-choice survey in which they were to suppose that they would develop AD in the future without medication. Choice tasks presented the option of no medication or a hypothetical AD treatment that would extend time with normal memory or time with cognitive impairment prior to developing AD. 12 or 16-year timeframes were used. Treatment benefits were combined with daily nausea and increased risks of disabling stroke and of death in the first year of treatment according to an experimental design. Choice data were analyzed using scale-adjusted latent-class analysis (LCA) in Latent GOLD. LCA revealed three distinct classes of respondents. Class one (42% of study sample) generally preferred medication, traded between all benefits and risks, was relatively younger, and was not likely to be a current AD caregiver. Class two (30%) preferred no medication, was more concerned about treatment risks than benefit, had no self-reported illnesses, and was somewhat more likely to be a current AD caregiver. Class three (28%) strongly preferred medication, was more concerned about treatment benefits than risks, was relatively older, and more likely to be a current AD caregiver. While most respondents (70%) were willing to accept treatment risks to reduce time with cognitive impairment or AD, 30% of respondents were risk averse with strong preferences for no AD treatment. LCA was useful for detecting heterogeneity of treatment preferences and may explain different rates of treatment uptake.

The objectives of this study were (1) to design and implement a stated-preference study to quantify the value of whole-genome sequencing (WGS) and (2) to evaluate the limitations of current stated-preference study-design methods for evaluating such complex health-care technologies. Following evaluations of early drafts in respondent interviews, the final study design incorporated solutions to three study-design challenges. First: multiple kinds and levels of uncertainty about the likelihood of numerous kinds of gene variants required simplifying the decision problem to evaluating a single gene variant with associated health consequences of varying severity. Second: the probability of getting WGS information about elevated health risks for which there are possible risk-reducing interventions, but have their own uncertain effects, required a simplified discrete-choice experiment (DCE) for the assumed health problem using surgery, medication, and watchful-waiting labeled alternatives. Third: valuing multiple kinds of test results required constructing contingent-valuation questions for a report containing only actionable information versus a report containing both actionable and non-actionable information. The online survey was administered to a US general-population sample of 410 respondents. Differences in DCE preferences for watchful waiting versus surgery or medication for a 20% chance of mild symptoms and a 60% chance of severe symptoms were statistically significant. A majority of respondents had no interest in non-actionable genomic information (55%, 95%CI: 50- 60%). Respondents with a positive value of information were willing to pay $299 (SD: $86, p<0.01) for actionable findings and $180 (SD: $83, p<0.05) for non-actionable findings. Our study complies with the ISPOR checklist for good stated-preference research practices and the results have reasonable face validity. However, necessary simplifications of the problem that departed substantially more than usual from the actual decision context suggest that current stated-reference methods have significant limitations for quantifying policy-relevant preferences for such complex technologies as WGS.

This study quantified patients’ stated benefit-risk tradeoff preferences for reduced severity and/or amount of psoriasis plaques relative to risk of symptomatic tuberculosis and risk of dying from serious infection with injected biologic medications. People with psoriasis were recruited to take the online survey via email invitation from the National Psoriasis Foundation. Respondents were randomly assigned to one of 8 survey versions containing 5 trade-off questions from the 40 questions in the D-efficient experimental design. Choice data were analyzed using a combination of latent-class analysis to identify groups with systematically different preferences and random-parameters logit (RPL) to account for unobserved taste heterogeneity within groups. 927 respondents provided useable survey data. RPL models on the full study cohort resulted in disordered preference weights indicating more severe conditions were preferred to milder conditions. Latent-class models identified three groups of respondents with distinctly different benefit-risk tradeoff preferences. Class membership was correlated with uneven assignment to the 8 survey versions. Controlling for respondents who always picked the better level of an attribute and weighting the data by the inverse of the proportion assigned to each survey version resulted in generally well-ordered categorical preference weights in the RPL models. Respondents had strong preferences for eliminating even small amounts of psoriasis. The 72% of respondents who accepted some benefit-risk tradeoffs were willing, on average, to accept a 10-year infection-related mortality risk of 3.2% [5.9%, 0.4%] to reduce the area of mild patches from 25% to 1% of body surface area or 3.5% [5.7%, 1.2%] to reduce the area of mild patches from 1% to clear. Uneven allocation across experimental design versions influenced the incidence of different choice patterns that latent-class models attributed to taste heterogeneity. Without correction, standard RPL modeling would have produced disordered preferences and biased estimates of acceptable benefit-risk tradeoffs.

BACKGROUND:Plaque psoriasis can have a significant negative effect on patients' quality of life, and treatments can result in serious toxicities. Although there have been several studies of patients' and physicians' relative preferences for the benefits and risks of psoriasis treatments, it is unclear how and whether patients' and physicians' preferences for the outcomes of psoriasis treatments differ.OBJECTIVES:To quantify patient and dermatologist preferences for improvements in psoriasis symptoms and for increases in the risk of treatment-related serious adverse events.METHODS:Members of the U.K. Psoriasis Association and U.K. dermatologists with experience prescribing biologics completed a web-enabled discrete-choice experiment survey in which they evaluated efficacy and safety features of biological treatments for psoriasis. Choices between hypothetical treatment options were used to estimate preference weights indicating respondents' relative trade-off preferences among treatment outcomes. These outcomes included improvements in the severity and coverage of psoriatic plaques and treatment-related risks of tuberculosis, serious infections and lymphoma. Preference estimates were used to derive the maximum level of side-effect risks that respondents would accept for improvements in psoriasis symptoms.RESULTS:Respondents' tolerance for side-effect risks varied with side-effect severity and location of plaques, and risk tolerance for serious side-effects was greater for patients than for dermatologists.CONCLUSIONS:Estimates of patients' risk tolerance for serious side-effects indicate that patients valued psoriasis symptom control highly and suggest that psoriasis symptoms have a significant effect on patients' quality of life. In light of research showing increased treatment satisfaction and improved treatment adherence among patients who receive therapies that are consistent with their preferences, our findings suggest that greater communication between dermatologists and patients about risk tolerance could help improve patient care.

Despite the existence of standardized medical criteria, clinical diagnosis of Alzheimer’s disease remains difficult. Lack of diagnostic certainty or possible distress related to positive results could limit application of new testing technologies. Independent of the therapeutic value of prevention or cure, however, diagnostic information could have value in informing contingency planning or have intrinsic value: the value of “just knowing”. This paper aims to quantify respondents’ preferences for obtaining AD diagnostic tests and to estimate the perceived value of AD test information. Discrete-choice experiment (DCE) and contingent-valuation (CV) questions were administered to N=1301 respondents aged 60 years or older in Germany and the United Kingdom. 12 pairs of virtual AD diagnostic tests were presented (defined by test type, test precision defined by false-positive or false-negative test results, and test cost) and a no-test alternative. DCE questions were based on a predetermined experimental design. A double-bounded, dichotomous-choice CV question was used to further elicit willingness to take an AD test and pay for it. Choice data of respondents interested in taking a test were analyzed using random-parameters logit. A probit model characterized respondents who were not willing to take a test. Most respondents in both countries had a positive value for AD diagnostic test information. Most respondents who indicated an interest in testing in Germany (N=631) and the UK (N=670) preferred brain imaging without radioactive marker, followed by brain imaging with radioactive marker and then spinal tap. Diagnostic tests with better precision were preferred. German respondents had relatively greater heterogeneous preferences and lower money-equivalent values for test features compared to UK respondents. Respondents preferred less invasive diagnostic procedures and tests with higher accuracy. Diagnostic test accuracy was more important than test type. Respondents expressed a willingness to pay up to €700 to receive a less invasive test with the highest accuracy.

OBJECTIVE The objectives were to quantify psychiatrists' judgments of the benefits and risks of antipsychotic treatments of patients with schizophrenia and to evaluate how patient adherence history affects these judgments. METHODS Weights assigned by respondents to risks, benefits, and alternative drug formulations in the treatment of schizophrenia were assessed via a Web-based survey by using a discrete-choice experiment. Respondents in the United States and the United Kingdom chose among alternative scenarios characterized by various levels of improvement in positive symptoms, negative symptoms, social functioning, weight gain, extrapyramidal symptoms (EPS), hyperprolactinemia, and hyperglycemia and by formulation. The effect of patient adherence history on respondents' judgments was also assessed. Random-parameters logit and bivariate probit models were estimated. RESULTS The sample included 394 psychiatrists. Improvement in positive symptoms from "no improvement" to "very much improved" was the most preferred outcome over the range of improvements included and was assigned a relative importance score of 10. Other outcomes, in decreasing order of importance, were improvement in negative symptoms from "no improvement" to "very much improved" (5.2; 95% confidence interval [CI]=4.2-6.2), social functioning from "severe problems" to "mild problems" (4.6, CI=3.8-5.4), no hyperglycemia (1.9, CI=1.5-2.4), <15% weight gain (1.5, CI=.9-2.0), no hyperprolactinemia (1.3, CI=.8-1.6), and no EPS (1.1, CI=.7-1.5). As adherence decreased, formulation became more important than modest efficacy changes and injections were preferred to daily pills (p<.05). CONCLUSIONS Psychiatrists favored treatments that primarily improve positive symptoms. Choice of formulation became more important as likely adherence declined.

AIMS:The aim of the study was to quantify patient preferences for outcomes associated with oral antidiabetic medications (OAMs) in Sweden and Germany through a discrete-choice experiment.METHODS:Adults taking OAMs who had a self-reported physician's diagnosis of type 2 diabetes mellitus (T2DM) made a series of nine choices between pairs of hypothetical profiles. Each profile had a predefined range of attributes: blood glucose control, frequency of mild-to-moderate hypoglycaemia, annual severe hypoglycaemic events, annual weight gain, pill burden and frequency of administration, and cost. Choice questions were based on an experimental design with known statistical properties. Bivariate probit analysis estimated the probabilities of choice of medication administration from patient characteristics and, conditional on that choice, preferences for treatment outcomes.RESULTS:The final sample consisted of 188 Swedish and 195 German patients. For both countries, weight gain was the most important attribute, followed by blood glucose control. Avoiding a 5-kg weight gain was 1.5 times more important in Sweden and 2.3 times more important in Germany than achieving moderate blood glucose control, thereby, suggesting that blood glucose control is relatively more important to Swedish than to German patients. Least important outcomes were the number of daily pills (Sweden) and frequency of mild-to-moderate hypoglycaemia (Germany).CONCLUSION:Patients in both Sweden and Germany preferred OAMs not associated with weight gain.