ORAL ABSTRACTS: WEIGHING THE RISKS OF INTERVENTION: PATIENT PERCEPTIONS AND INFORMED CHOICE
Anne M. Stiggelbout, PhD
Leiden University Medical Center
Purpose: To evaluate the quality of informed decision-making (IDM) and the degree uncertainty is acknowledged in patient-surgeon decisions about high-risk surgery.
Methods: This prospective, multi-institutional study evaluated 90 preoperative patient-surgeon discussions prior to high-risk cardiothoracic, vascular, oncologic, and neurosurgical operations in the U.S. and Canada. Consultations were audiotaped, transcribed verbatim, and coded using Braddock's model of 9 IDM elements (see Table). An interdisciplinary team of 5 researchers scored the transcripts (inter-rater reliability, κ=0.89).
Results: Surgeons commonly discussed the risks and benefits of surgery (92%), nature of the decision (60%), and alternatives (46%). Risk communication was most often numeric and narrative (33%) or numeric only (33%) and included mortality in 58%. Surgeons uncommonly assessed patients' desire for others' input (1%), preferences (12%), or the impact of surgery on daily life (14%). The Table lists the extent to which surgeons partially or fully discussed each IDM element. Only 26% of discussions met Braddock's minimum criteria for IDM, which requires reviewing the patient's role or preference and the nature of the decision. Uncertainty about the decision to have high-risk surgery or alternative treatment was fully discussed in 19% of consultations and partially discussed in 10%. Surgeons also conveyed uncertainty about the underlying disease (33%), post-surgical outcomes (24%), and other aspects of care (28%). If only one treatment option was offered, IDM was less extensive (mean IDM elements discussed: 7.0±3.2 vs. 5.3 ±2.9, p=0.01), and the greatest decrease was in uncertainty (49% vs. 15%, p<0.001).
Assessment of decision-making demonstrated that patients and surgeons frequently made the decision together (68%). Surgeons were much more likely to make recommendations (54%) than solicit patient's preferences (12%), though 20% of patients or their families offered treatment preferences without solicitation. In 9% of consultations, patients were the primary decision maker; and IDM was more extensive than when surgeons were the dominant decision maker (7.0±3.7 vs. 4.6±2.0, p=0.06). Without surgeon prompt, 30% of patients and/or families initiated discussion of an IDM element, and 26% asked unsolicited questions about an IDM element.
Conclusions: In patient-surgeon decisions about high-risk surgery, IDM quality is highest in areas that overlap with informed consent—risks, benefits, and alternatives. However, the extent to which patients' preferences, role, and uncertainty are incorporated in the decision-making process appears to need improvement.
Method: A discrete choice experiment (DCE) was used to value preferences for attributes of bariatric surgery procedures. Attributes of the DCE were identified through 12 patient focus groups (n=93) and an expert stakeholder panel of bariatric program coordinators who were also previous bariatric surgery patients (n=12). Using thematic analysis, 9 attributes were identified as being most important to the consideration of weight loss treatment options, which included bariatric surgery and medical weight loss. Attributes were: how treatment works, extensiveness, years treatment has been available, weight loss, resolution of medical conditions, risk of complications, side effects, diet changes, and out of pocket costs. Profiles of hypothetical weight loss or bariatric surgery options were generated using a fractional factorial design. Respondents were asked to select which option they would choose from pairs of hypothetical treatment profiles or an opt-out option. Respondents were recruited from bariatric surgery information sessions in 2015. An effects coded conditional logit model was used to assess the relative importance of each attribute on choices. Latent class analysis was used to identify patient subgroups with distinct preferences.
Result: Respondents reflected the general population of bariatric surgery candidates with most between the ages of 30-59 and 80% women (n=815). Higher total weight loss, lower initial out-of-pocket cost, and resolution of medical conditions were strongly associated with a higher likelihood of selecting a profile. How treatment works, risk of complications, side effects, and dietary changes were of lower significance. Latent class analysis revealed a group of respondents that was more sensitive to out-of-pocket costs and second group that was more sensitive to the weight and health effects of a procedure.
Conclusion: Prospective patients placed a high value on total weight loss, low initial out of pocket costs, and a resolution of their medical conditions, but did not place much consideration on risk of complications or side effects of available treatments. One segment of patients was very sensitive to the out-of-pocket costs associated with available treatment options.
Purpose: To examine how much variations in patient treatment burden would influence optimal use of antihypertensive medications and how much over- and under-treatment can result from clinicians misunderstanding their patients' treatment burden.
Method: We developed a Markov Decision Process (MDP) model to estimate the difference in expected QALYs (over a 10-year planning horizon) if clinicians used the true patient burden to guide blood pressure treatment, compared to the expected QALYs if the clinician's perceived patient burden diverged from the true patient burden. The MDP model makes decisions that maximize expected QALYs for each patient using the clinician's perceived treatment burden, thus estimating the best care that the clinician could provide if the only limitation is their uncertainty about patient (true) treatment burden. The models were calibrated using CVD risk factors derived from NHANES as well as other published data. Untreated event rates were estimated using the Framingham score. Sensitivity analyses was performed to examine: (1) the definition of mistreatment, (2) the discount factor used, (3) the mortality scaling factor, (4) variability around the perceived and true burden, and (5) patient adherence.
Result: Fairly small differences in true patient burden from blood pressure treatment alter the number of blood pressure medications that should be recommended and alters treatment's potential benefit dramatically. We also found that a clinician misunderstanding the patient's burden could lead to almost 30% of patients being treated inappropriately. Most dramatically, if a clinician's underestimation of the patient's true burden makes that patient less likely to take a medication, this would amount to losses of up to 85.3 QALYs per 1000 people.
Conclusion: Treatment burden varies between patients based on side effects, costs, and patient opinion. Current clinical practice fails to gather information that explicitly understands this variation, and clinical practice guidelines fail to account for it. We found that clinical decisions that fail to account for patient treatment burden can mistreat a very large proportion of the public. Successful treatment selection is closely dependent on a clinicians' ability to accurately gauge a patient's treatment burden.
Method: A sample of 508 US adults ages 40-75 was recruited from a national online survey panel. The sample was roughly representative of the US population on gender (52% male), ethnicity (28% non-White) and education (33% high school diploma or less). Respondents were asked their preferences for the presentation of prescription drug risk information. In addition, an experiment examined which risk information actually influenced patients’ intentions to use a prescription drug. Respondents were randomly assigned to different versions of a prescription drug description, which varied on side-effect severity, side-effect likelihood, and seriousness of the disease being treated. Afterwards they reported their intentions to use or recommend the drug.
Result: Most respondents stated they wanted to be told about all possible side effects. However, experimental findings indicated that the decision impact of risk information varied considerably depending on the risk’s severity and likelihood, and the seriousness of the disease being treated. The possibility of a mild side effect (mild headache) had no impact on drug-use intentions, regardless of likelihood (1 in 10,000 patients vs. 1011 in 10,000 patients) or the severity of the disease being treated (eczema vs. Alzheimer’s). However, the decision impact of a potential severe side effect (internal bleeding) was more complex. For a drug treating a relatively mild disease (eczema), the mere possibility of a severe side effect sharply reduced drug-usage intention, even if this side effect was extremely unlikely. However, in treating a debilitating disease (Alzheimer’s), patients were willing to tolerate a potentially severe side effect if it was extremely unlikely (1 in 10,000).
Conclusion: The recommendation that drug risk disclosures focus on the “most important” risks seems reasonable in theory, but may be challenging to implement, since patient perceptions of risk importance involve complex interactions of perceived severity, perceived likelihood, and the disease being treated.
Purpose: Understanding patients' willingness to accept treatment-related risks of serious adverse events is important in selecting optimal treatments. However, common practice of reporting average estimates of risk tolerance could lead to incorrect inferences for many patients. This study assessed heterogeneity in patients' risk tolerance for serious infection, cancer, and surgery with Crohn's disease (CD) medication treatments.
Methods: We administered a discrete-choice experiment survey to patients enrolled in an online panel of CD patients. Participants completed 8 choice questions that required comparing two constructed medical therapies. Each treatment option was characterized by differing levels of time with active disease symptoms, severity of symptoms, duration of therapy with corticosteroids, and risks of serious infection, cancer, and requiring surgery. Scale-adjusted latent-class choice models identified groups of patients with similar treatment-outcome preferences and risk tolerances.
Results: The analysis identified three distinct subgroups in the sample of 811 participants characterized by: 1) those disproportionately concerned with duration of active symptoms (61%), 2) those disproportionately concerned with steroids (25%), or 3) those disproportionately concerned with treatment-related risks (14%). (See Figure.) Class membership was significantly correlated with severity of current and most-recent relapse symptoms, experience with CD complications, and having a family member with cancer. The more risk-tolerant respondents more often were under age 44 (p<0.01) and were more likely to be male (p<0.01). For a treatment that would avoid 12 months of moderate CD symptoms, patients in the 3 classes would accept serious-infection risks of more than the maximum 30% evaluated in the study design, 12%, and 16%, respectively. Conversely, for a treatment with a 5% increased risk of requiring surgery, patients in the 3 classes would require reductions in moderate symptoms of 1.2 months, 4.5 months, and more than the maximum 12 months evaluated in the survey design.
Conclusions: There is substantial preference heterogeneity among patients with CD for medications that vary in efficacy and potential harms. Avoiding the ecological fallacy of assuming all patients have mean-value preferences could result in significant differences in the conclusions of studies intended to support clinical and regulatory decision making.
Purpose: Identifying the most influential factors in the decision to vaccinate may help policy makers design more effective programs to increase vaccination rates. The purpose of this study was to identify the relative importance of attributes that influence vaccine decision-making among adults, adolescents, and parents of adolescents in the United States using conjoint analysis.
Method: Twelve attributes related to vaccine uptake were selected for inclusion in a conjoint analysis based on findings from four focus groups and a literature review (Table 1). Each attribute was assigned 2-8 possible values, or levels. Survey respondents were given a profile of a hypothetical vaccine based on selected attribute levels and asked whether they would accept (yes/no) the vaccine for themselves (adults and adolescents) or their child (parents). Respondents could be randomly assigned to 13 different versions of the survey. Each version contained 5 vaccine profiles, generated using a fractional factorial design, with 8 alternating attributes in each. The survey was fielded to a sample of adolescents (ages 13-17; n=1516), parents (n=1030), and non-parent adults (n=652) in November 2015 using a national probability-based online research panel. Logistic regression with generalized estimating equations and effect coding was used to estimate the relative value of each attribute.
Result: The survey was completed by 316 adolescents, 339 parents of adolescents, and 334 adults (31% overall response rate). Vaccine acceptance varied, with 55% of adolescents, 62% of parents, and 50% of adults willing to accept a vaccine based on the profile presented. Adolescents were most likely to accept vaccines with high effectiveness (OR=2.43, 95% CI: 1.63-3.62), strong primary care physician (PCP) recommendation (OR=2.03, 95% CI: 1.57-2.62), and moderate cost (OR= 1.46, 95% CI: 1.09-1.97, for $75 cost after insurance). Parents and adults had similar values, with both also indicating that high vaccine effectiveness was the most important attribute, but adults considered low cost (OR= 1.86, 95% CI: 1.36-2.55 for $10 cost after insurance) to be more influential than a strong PCP recommendation (OR=1.52, 95% CI: 1.20-1.94).
Conclusion: Adolescents, parents and adults display consistent preferences with regards to vaccine decisions and are most sensitive to the vaccine effectiveness, followed by PCP recommendations and cost. Health policies that focus on patient–provider communication and vaccine affordability may be effective levers to increase vaccine uptake.