* Candidate for the Lee B. Lusted Student Prize Competition
Purpose: In health state valuations the public fails to anticipate adaptation-- they neglect to consider their capability of making the best of a situation. But do people who actually have a new disability consider adaptation in making forecasts?
Method: Patients who had recently had colostomy surgery (n=67) and patients who had recently had amputation surgery (n=56) received a survey at three different time points. The surveys consisted of measurements of physical and psychological well-being (quality of life, satisfaction with life, general health, physical functioning, activities, activities outside, calm/peacefull, energy, downhearted, embarrassed). Some items asked patients to predict how they would feel at the next survey time point—and to remember how they felt at the previous survey time point.
Result: From the predictions and actual valuations of the different well-being scales we find a clear pattern. Patients expected to improve on most well-being domains over time. In general, they overpredicted the amount of improvement. For example, colostomy patients at baseline predicted that their general health at the second wave would be significantly higher than their current general health at baseline ((M = 3.13 vs. 2.78, p < .05), and this prediction was significantly higher than their actual reported health at the second wave (M = 3.13 vs. M = 2.83, p < .05). Looking at recall, we find a similar pattern. Patients tend to think they are better off now than they were previously, and they underestimate their well being at the prior time points.
Conclusion: Patients who are in a situation to which they can adapt are overly optimistic in terms of predictions, and overestimate adaptation in their recollections. This is in contrast to what might be expected from the past literature on affective forecasting and may have implications for how treatment decisions are discussed.
Purpose: Cesarean deliveries account for 31.8% of births in the US. “Cesarean delivery on maternal request” (CDMR) has been suggested as contributing to this increase. However, little is known about how women value alternative delivery approaches and their potential outcomes, and to what extent they would choose to have a planned cesarean delivery if offered. We sought to assess pregnant women's mode of delivery preferences and to identify sociodemographic determinants of these preferences.
Method: Women receiving prenatal care at the University of California, San Francisco, or San Francisco General Hospital were invited to participate in an interview at 22-35 weeks gestation. Time tradeoff (TTO) and standard gamble (SG) utilities, with attempted vaginal delivery ending in cesarean or a cesarean with complications as bottom anchors, for 8 scenarios associated with either a planned vaginal or planned cesarean delivery were measured using a computerized preference elicitation tool (ELICIT). A demographic/attitudinal questionnaire also was administered.
Result: Of the 81 participants, 57% were multiparous and 6% had a prior cesarean delivery. 95% preferred vaginal delivery and, on average, were willing to accept a 74% chance of their planned vaginal delivery ending in a cesarean before switching their preference to a planned cesarean delivery. In a regression model adjusted for age, parity, and race/ethnicity, women with at least a college degree were more likely than those without a degree to indicate that they would attempt a vaginal delivery even with a 0% chance of success (SG=0, adjusted OR 14.1, 95% CI 2.1-94.7). In other words, these women highly valued experiencing labor even if it would not culminate in a vaginal delivery. Median SG and TTO scores for the remaining delivery scenarios ranged from 0.19-0.75 and 0.97-0.99, respectively.
Conclusion: The vast majority of participants preferred vaginal delivery and would attempt this approach even if the chance of success was low. The strength of this preference varied depending on maternal education and the specific delivery scenario being assessed. Few women would accept a reduced life expectancy to achieve their preferred mode of delivery. In this population, CDMR is unlikely to be a primary cause of the increasing cesarean delivery rate. Further studies in other geographic areas are needed to shed further light on the role of preferences in mode of delivery decisions.
Purpose: There is rising interest in estimating health-state valuations from ordinal data, although key methodological challenges remain. Recent studies suggest that combining information elicited using ordinal and cardinal techniques may optimize tradeoffs between simpler data collection and valid estimation. Analytic models for ordinal data typically assume constant variance across states and unbiased measurement. This study used simulation and analysis of a large empirical data set to quantify the potential importance of three departures from these standard assumptions.
Method: We applied a probit model that combines information from paired comparisons of heath states with discrete choice time trade-off (TTO) responses. The model assumes that values for health states on a (0,1) scale are compressed at either end of the scale, such that the log-odds of these values are normally distributed with constant variance. We generated simulated datasets that deviated from these assumptions and then tested the robustness of the analytic models to these deviations. Specifically, we simulated (1) heteroskedastic variance (in log-odds space) across health states; (2) bias due to non-detectable differences between similar health states; and (3) strict preference for longevity over perfect health in TTO responses among some proportion of respondents (“never traders”). In light of the results of the simulation study we tested alternative modeling strategies in a large, multi-country empirical database on health valuations.
Result: For simulations in which all assumptions were met (and given a typical sample size for health valuation) the model accurately estimated health-state values, with a mean relative error (MRE) of 6% for predicted vs. true values. If variances for intermediate states were higher than for extreme states (by a factor of 4 in log-odds space), the model still performed well (MRE=10%). Results were more sensitive to the other two departures. For example, when 15% of paired comparisons produced random errors due to non-detectable differences, MRE rose to 25%, and when 20% of respondents were “never traders”, MRE was 23%. Addressing the problems of non-detectable differences and “never traders” in empirical data demanded both larger numbers of TTO responses and augmented analytic models.
Conclusion: Ordinal approaches to health-state valuation are attractive alternatives to existing metric methods, but further work is needed in identifying the most appropriate methods for data collection and analysis.
Purpose: Prevention guidelines indicate that large numbers of middle aged and older people should use statins. In practice, many fewer are prescribed such drugs, and considerable proportions discontinue their treatment. The purpose of this study was to identify factors that may influence people’s decisions regarding a cardiovascular prevention drug.
Method: A representative sample of 4,000 individuals aged 40-69 in Odense, Denmark, was randomly selected and invited for an interview. 1,491 (37%) were successfully interviewed, and 1,169 interviews were used in this study. Subjects were randomised into 16 different groups, comprising different baseline risk levels (5% or 15% 10-year risk of fatal heart attack) and effectiveness groups (2%, 4%, 5%, 10% in terms of absolute risk reduction). Interviewees were asked to imagine themselves at increased risk of cardiovascular disease and for that reason offered treatment with a hypothetical drug. They received comprehensive information about the effectiveness in terms of absolute risk reduction, relative risk reduction, number needed to treat, and life extension. Subsequently, they were asked whether they would or would not consent to therapy. Finally, they were asked about the reasons for consenting or not consenting.
Result: For absolute risk reductions of 2%, 4%, 5% and 10%, the proportion of subjects accepting treatment were 57%, 69%, 68% and 73%, respectively. Among those who consented to therapy, 45% said it was because of their health, 32% because of family considerations, and 17% because of confidence in the doctor. Among those who rejected therapy, preference for life-style changes (56%), fear of side-effects (19%), and low effectiveness (13%) were the most frequently stated reasons. Reasons were independent of socio-demographic characteristics and presentation of effectiveness information.
Conclusion: The level of health benefit seems to have a moderate influence on people’s decisions about preventive drugs while important personal and inter-personal aspects, e.g. family situation, availability of non-medical alternatives, and trust in the doctor were reported as influencing decisions. The findings suggest that physicians may do well in discussing these reasons for treatment decisions with their patients in order to make optimal decisions.
Purpose: Lottery-based incentives are being used increasingly to augment response in clinician surveys. Although lotteries may be more convenient and less costly per response than fixed payments, their effects on response rate are uncertain.
Methods: We conducted three randomized trials of fixed payments and actuarially equivalent financial lotteries. In Trial 1, we administered a web-based questionnaire regarding deceased organ donor management to U.S. critical care physicians. Physicians were randomly assigned to a “jackpot lottery” (60%) in which responding would provide a 1 in 1000 chance of winning $5000, to a “high-probability lottery” (20%) in which responding would provide a 1 in 50 chance of winning $250, and to no financial incentive (20%). In Trial 2, we administered identical questionnaires via post to U.S. critical care nurses. Nurses were randomly assigned to a lottery (75%) with a 1 in 400 chance of winning $2000 or to receive an unconditional fixed payment – a $5 bill in the initial outgoing envelope (25%). In Trial 3, we administered a web-based questionnaire regarding consent for blood transfusion to resident physicians. Residents were randomly assigned to a lottery (50%) with a 1 in 250 chance of winning $2500 or to receive a conditional fixed payment – a promise that $10 would be mailed to them after completing the questionnaire (50%).
Results: Among 2,206 eligible physicians in Trial 1, response rates were 30.5% in the jackpot lottery, 31.2% in the high-probability lottery, and 32.4% in the no-incentive group (all p > 0.30). Among 988 eligible nurses in Trial 2, response rates were 39.4% in the lottery group and 58.8% in the unconditional fixed payment group (p < 0.001). Among 758 eligible residents in Trial 3, response rates were 51.1% in the lottery group and 55.8% in the conditional fixed payment group (p = 0.20).
Conclusion: Lotteries appear to have little utility in clinician surveys: they do not improve response rate compared with no incentives and they produce lower response rates than actuarially equivalent fixed payments. The finding that unconditional fixed payments (those provided up front) are superior to lotteries, whereas conditional fixed payments (those provided following response) are not suggests that the ability to introduce a duty of reciprocity, rather than merely guaranteeing reimbursement, is essential for fixed payments to work.
Purpose: Proposals to increase the supply of transplantable kidneys through government-regulated payments to living kidney donors have not been seriously considered by policymakers due to ethical concerns. We sought to determine the extents to which the three main ethical concerns with paying kidney donors might manifest if regulated payments were legalized.
Methods: We recruited a representative sample of Philadelphia County residents by interviewing consecutively encountered regional rail passengers. We used conjoint analysis to elicit participants’ willingness to donate a kidney while systematically varying (1) the amount they would be paid, (2) their risk of subsequently developing kidney failure themselves, and (3) who would receive their kidney. We determined whether payment represents an “undue inducement” by evaluating participants’ sensitivity to risk as a function of the payment offered; whether payment represents an “unjust inducement” by evaluating participants’ sensitivity to payment as a function of their annual income; and whether introducing payment would reduce altruistic donations by comparing the proportions of participants who would donate altruistically (for free) when this condition was presented up front versus after participants were exposed to payments.
Results: Among 342 encountered rail passengers, 260 completed the questionnaire (response rate = 76.0%), and 222 were medically eligible to donate. Clustered logistic regression revealed that donation rates increased significantly as risk of developing kidney failure decreased, as the payment offered increased, and when the kidney recipient was a family member rather than the next patient on the public waiting list (p < 0.01 for each). Despite greater than 90% power to detect statistical interactions, none was found between payment and risk (p = 0.80) or between payment and participants’ income (p = 0.20), suggesting that payment is neither an undue nor an unjust inducement, respectively. Alerting participants to the possibility of payment did not alter their willingness to donate for altruistic reasons (p = 0.77).
Conclusion: Theoretical concerns with paying people for living kidney donation are not corroborated by empirical evidence. We cannot recommend establishing a kidney market based on these results because choices revealed in hypothetical scenarios may not reflect real-world behaviors. However, these results suggest that it is time to conduct a real-world test of regulated payments to determine whether they may reduce the kidney shortage ethically.