Patient Preferences for Bariatric Surgery: Findings From a Survey Using Discrete Choice Experiment Methodology

Key Points

Question

What value do patients eligible for bariatric surgery place on the risks and benefits of the characteristics of different procedures?

Findings

In a discrete choice experiment with 815 adults considering bariatric surgery, choice of procedure was most strongly based on cost, total weight loss, and resolution of medical conditions.

Meaning

Results of our study suggest that a hierarchy of procedure characteristics can be used by physicians to help patients select their preferred weight loss option.

This survey study identifies patient self-reported preferences for risks, benefits, and other attributes of treatment options available to individuals who are candidates for bariatric surgery.

Abstract

Importance

Surgical options for weight loss vary considerably in risks and benefits, but the relative importance of procedure-associated characteristics in patient decision making is largely unknown.

Objective

To identify patient preferences for risks, benefits, and other attributes of treatment options available to individuals who are candidates for bariatric surgery.

Design, Setting, and Participants

This discrete choice experiment of weight loss procedures was performed as an internet-based survey administered to patients recruited from bariatric surgery information sessions in the State of Michigan. Each procedure was described by the following set of attributes: (1) treatment method, (2) recovery and reversibility, (3) time that treatment has been available, (4) expected weight loss, (5) effect on other medical conditions, (6) risk of complication, (7) adverse effects, (8) changes to diet, and (9) out-of-pocket costs. Participants chose between surgical profiles by comparing attributes. Survey data were collected from May 1, 2015, through January 30, 2016, and analyzed from February 1 to June 30, 2016.

Main Outcomes and Measures

Estimated relative value of risks and benefits for leading weight-loss surgical options and marginal willingness to pay for procedure attributes. A latent class analysis identified respondent subgroups.

Results

Among the 815 respondents (79.9% women; mean [SD] age, 44.5 [12.0] years), profiles of hypothetical procedures that included resolution of medical conditions (coefficient for full resolution, 0.229 [95% CI, 0.177 to 0.280; P < .001]; coefficient for no resolution, −0.207 [95% CI, −0.254 to −0.159; P < .001]), higher total weight loss (coefficient for each additional 20% loss, 0.185 [95% CI, 0.166 to 0.205; P < .001]), and lower out-of-pocket costs (coefficient for each additional $1000, −0.034 [95% CI, −0.042 to −0.025; P < .001]) were most likely to be selected. Younger respondents were more likely than older respondents to choose treatments with higher weight loss (coefficient for loss of 80% excess weight 0.543 [95% CI, 0.435-0.651] vs 0.397 [95% CI, 0.315-0.482]) and were more sensitive to out-of-pocket costs (coefficient for $100 out-of-pocket costs, 0.346 [95% CI, 0.221-0.470] vs 0.262 [95% CI, 0.174 to 0.350]; coefficient for $15 000 in out-of-pocket costs, −0.768 [95% CI, −0.938 to −0.598] vs −0.384 [95% CI, −0.500 to −0.268]). Marginal willingness to pay indicated respondents would pay $5470 for losing each additional 20% of excess body weight and $12 843 for resolution of existing medical conditions, the most desired procedure attributes. Latent class analysis identified the following 3 unobserved subgroups: cost-sensitive (most concerned with costs); benefit-focused (most concerned with excess weight loss and resolution of medical conditions); and procedure-focused (most concerned with how the treatment itself worked, including recovery and reversibility).

Conclusions and Relevance

Candidates for bariatric surgery identified costs, expected weight loss, and resolution of medical conditions as the most important characteristics of weight loss surgery decisions. Other information, such as risk of complications and adverse effects, were important to patients but less so.

Introduction

Bariatric surgery is a safe and effective intervention for the treatment of morbid obesity and its associated comorbidities. A growing literature is available for patient preferences for weight loss treatments, including bariatric surgery.^1,2 Most existing studies have approached the question of patient preference by taking a procedure-centered focus in which study participants are asked reasons for and against particular procedures.

However, little is known about how a patient’s preferences concerning the medical and financial nuances of a procedure affect his or her decision to pursue one operation over another. The variation in procedure outcomes, complications, extensiveness of the procedure, weight loss mechanism, and reversibility make the choice of weight loss method a clear example of a preference-sensitive decision. Although existing studies effectively identify which procedure characteristics motivate a particular procedure’s selection, no study has adequately provided a quantitative assessment of trade-offs among characteristics.

In this context, our study makes 2 important contributions to the existing literature. First, it provides a quantitative valuation of the relative importance of procedure characteristics in treatment decisions from the patient perspective using conjoint analysis, a survey-based method designed for evaluating trade-offs across characteristics or attributes. Second, this study complements existing studies on patient preferences for bariatric surgery by using a different starting point. Instead of asking patients about the reasons for or against particular procedures, we ask patients about key procedure characteristics.

Methods

We conducted an online survey to understand patient bariatric surgery preferences. Candidates were recruited from bariatric surgery information sessions throughout the State of Michigan, a state with a slightly greater preference for sleeve gastrectomy (SG)³ when compared with the national average.⁴ The institutional review board of the University of Michigan, Ann Arbor, approved all study procedures, and all participants provided written informed consent.

Conjoint Analysis Methodology

This study used conjoint analysis, a survey-based approach for measuring preferences.⁵ The premise of conjoint analysis, originally developed for the marketing and transportation sectors, is that any good or service can be described by its characteristics or attributes, and the extent to which an individual values a particular attribute can be quantified and compared with other attributes.⁶ This premise is especially useful when evaluating trade-offs between competing options. The methodologic strength of conjoint analysis is offering a quantitative assessment of the most valued attributes of a good or service. At the same time, the method is not designed to easily compare existing products as a whole. Therefore, the results allow us to compare key aspects of surgical procedures but not to compare the procedures themselves.

Studies using conjoint analysis are increasingly used in health services research to measure preferences for health interventions and public health programs.⁷ This method is often used to establish patient, physician, or public preferences for medical care, such as determining which factors parents prioritize when selecting a pediatric medical home.⁸ Surgical applications have been used to demonstrate that patients undergoing cataract surgery in the United Kingdom prioritized damage to sight and wait time over surgeon grade⁹ and that family physicians in Canada recommended lumbar spinal surgery based on different criteria than the surgeons themselves.¹⁰

For this study, we used a discrete choice experiment, a type of conjoint analysis that asks respondents to make a choice among 2 or more alternatives based on component characteristics of the alternatives.¹¹ Alternatives presented to respondents were hypothetical surgical and medical weight loss profiles. By varying the characteristics of competing profiles, this type of study determines the characteristics most important to a respondent’s choice.

Survey Development

To develop the experiment, we followed the recommendations of the International Society for Pharmacoeconomics and Outcomes Research checklist for conjoint analysis applications in health.¹² The first tasks include identifying the research question, determining the attributes and levels of the good or service, constructing the choice task, and determining the experimental design.

Key research questions were: (1) Which characteristics of bariatric surgical options are most important to potential patients? and (2) What is the relative value of their importance? An expert stakeholder panel consisting of 12 bariatric program coordinators who had also undergone bariatric surgery in the past was convened to develop a list of the most important attributes considered in choosing a bariatric surgical procedure. These attributes were further refined through 12 patient focus groups held in Detroit, Grand Rapids, and Traverse City, Michigan. Focus group participants (n = 93) had participated in a medical weight loss program or had laparoscopic-adjustable gastric banding, Roux-en-Y gastric bypass (RYGB), SG, or duodenal switch procedures. Focus group participants were given a $100 incentive.

Based on the focus groups, 9 attributes of bariatric surgical procedures were included in the final survey (eTable 1 in the Supplement). Levels for the weight loss attribute were expressed in terms of percentage of excess weight loss. The respondent-specific weight loss was calculated using the respondent’s reported body mass index and included in the profile descriptions to improve understanding of the weight loss levels. Attributes were defined to reflect a range that would encompass the 4 available bariatric surgical options and medical weight loss as well as include plausible attributes for procedures that may arise in the future.

Using a fractional factorial design,¹³ 64 full profiles of bariatric surgery procedures were generated. Respondents were randomized to 1 of 8 question blocks. A fractional factorial design is a subset of a full factorial design, chosen to pair each attribute with and against other attributes in the study.

Cognitive pretests were conducted to ensure face validity of the survey instrument and comprehensibility of the survey method. We conducted iterative pretests on paper (n = 7) and, after revisions, conducted additional pretests in an online format (n = 10). Respondents were provided pretest questions and were given the option to verbally share comments during the pretest.

The survey instrument contained an introductory section that explained the task and defined the attributes used in the survey. Respondents then answered 1 practice discrete choice question that was not scored and 8 scored discrete choice questions. Discrete choice questions presented 2 hypothetical bariatric surgery profiles. Respondents were asked to select which procedure they would prefer or indicate that they would not choose either procedure (Figure 1). An opt-out option was important to include because patients who initially would consider a surgical option for weight loss may ultimately decide against any available option based on the combination of attributes in the profiles. Finally, respondents provided demographic information and indicated if they found the discrete choice questions difficult to answer and, if so, why.

Figure 1. Sample Discrete Choice Task.

To convert pounds to kilograms, multiply by 0.45.

Study Population and Data Collection

Respondents were recruited at bariatric surgery information sessions. The survey was fielded from May 1, 2015, through January 30, 2016, and took approximately 15 minutes to complete. Respondents were given a $50 incentive.

Statistical Analyses

Data were analyzed from February 1 to June 30, 2016. The primary analysis used an effects-coded conditional logit adjusted for clustering at the respondent level.^6,14 Stratified analyses were conducted based on sex (male or female), income level (≤$25 000 or >$25 000), and age (<45 or ≥45 years). A secondary analysis was restricted to the 726 respondents (89.1%) who indicated they did not find the conjoint questions difficult to answer. All regression analyses were conducted using Stata software (version 13; StataCorp).

An additional analysis used a model in which out-of-pocket cost, risk of complications, total weight loss, and time the treatment has been available were all recoded as continuous variables. Recoding initial out-of-pocket costs as a continuous variable also allowed us to calculate marginal willingness to pay.

Additional analyses recoded certain variables as dichotomous instead of the effects coding. The extensiveness attribute was separated into 2 separate variables, with one based on weeks of recovery and the other based on reversibility. We combined years of treatment into 2 instead of 4 levels as newer (1 and 5 years) and established (10 and 20 years). We separated weight loss into 2 variables based on the amount lost during year 1 and the amount regained during year 2. We combined risk of complication into 2 instead of 4 levels as a low complication rate (0% and 5%) and a high complication rate (20% and 35%).

Finally, we conducted a latent class analysis to determine whether unobserved subgroups were driven by particular attributes more than other respondents.¹⁵ The cluster model of latent class analysis is conducted by introducing dummy variables into the conditional logit model based on the number of classes that may plausibly exist. A maximum likelihood estimation then determines the probability that a respondent would provide the observed responses, and the model that best categorizes the respondents is used. The latent class analysis was conducted using Latent GOLD (version 5.1; Statistical Innovations, Inc).

Results

Survey Respondents

A total of 815 people completed the discrete choice experiment (Table 1), making this, to our knowledge, the largest study on patient preferences for bariatric surgery to date. Respondents were much more likely to be female (651 [79.9%]) than male (164 [20.1%]). Six hundred twenty-five respondents (76.7%) were aged 30 to 59 years (mean [SD] age, 44.5 [12.0] years). The distribution of sex and age is typical of the population seeking bariatric surgery.¹⁶ The racial and ethnic backgrounds reflect Michigan’s general population, with most respondents identifying as white only (613 [75.2%]), and a large minority identifying as black or African American only (140 [17.2%]). A small number of respondents identified as other racial or ethnic minorities. Other characteristics, such as marital status, income, and educational level, were similar to those of the general population of the state. As expected, respondents were more likely to indicate having a medical condition (223 [27.4%] with diabetes; 386 [47.4%] with arthritis) or being disabled (128 [15.7%]) than the general population.

Table 1. Respondent Demographics.

Characteristic	No. (%) of Respondents (n = 815)
Age, ya
18-29	92 (11.3)
30-44	321 (39.4)
45-59	304 (37.4)
≥60	96 (11.8)
Sex
Female	651 (79.9)
Male	164 (20.1)
Race
Native American or Alaskan	4 (0.5)
Native American only	1 (0.1)
Asian	1 (0.1)
Asian only	0
Black or African American	153 (18.8)
African American only	140 (17.2)
Native Hawaiian or Pacific Islander	2 (0.2)
Native Hawaiian only	0
White	652 (80.0)
White only	613 (75.2)
Other	19 (2.3)
Other only	15 (1.8)
≥2 races	46 (5.6)
Ethnicity
Hispanic	35 (4.3)
Non-Hispanic	780 (95.7)
Highest educational attainment
<8th grade	4 (0.5)
Some high school	30 (3.7)
High school graduate	136 (16.7)
Some college	392 (48.1)
College graduate	165 (20.2)
Graduate school	88 (10.8)
Employment status
Not looking	16 (2.0)
Looking	21 (2.6)
Disabled	128 (15.7)
Part-time	78 (9.6)
Student	21 (2.6)
Retired	55 (6.7)
Homemaker	53 (6.5)
Full-time	422 (51.8)
Other	21 (2.6)
Household income, $
<10 000	112 (13.7)
10 000-24 999	159 (19.5)
25 000-44 999	174 (21.3)
45 000-74 999	177 (21.7)
75 000-100 000	89 (10.9)
>100 000	104 (12.8)
Marital status
Divorced or separated	157 (19.3)
Married	447 (54.8)
Single	195 (23.9)
Widowed	16 (2.0)
Medical conditions
Diabetes	223 (27.4)
Arthritis	386 (47.4)
Asthma	203 (24.9)
Smoker	76 (9.3)
Confidence in responsesb
Very confident	204 (25.0)
Somewhat confident	522 (64.0)
A little confident	52 (6.4)
Not at all confident	22 (2.7)

^aTwo respondents did not provide plausible dates of birth.

^bFifteen respondents did not complete this question.

Attributes

A coefficient value from a discrete choice experiment that is greater than 0 indicates that including that attribute level in a surgical profile makes it more likely that the profile will be selected. For example, the coefficient for loss of 80% of excess body weight is 0.460. This result means a surgical profile was more likely to be selected if it included that attribute level. A coefficient value less than 0 indicates a profile is less likely to be chosen when it included that attribute level. For example, the coefficient for loss of 20% of excess body weight is −0.473. This result means a surgical profile was less likely to be selected if it included that attribute level of loss of 20%.

The most significant attributes for respondent choice were resolution of medical conditions (coefficient value for no weight-associated conditions, 0.229; 95% CI, 0.177-0.280), total weight loss (coefficient value for loss of 80% of excess weight, 0.460; 95% CI, 0.373-0.546), and initial out-of-pocket costs (coefficient value for $100 out-of-pocket costs, 0.309; 95% CI, 0.211-0.406) (Figure 2 and Table 2). Resolution of sleep apnea, hypertension, high cholesterol levels, and diabetes made a profile more likely to be chosen, whereas a profile that failed to resolve any of those conditions made it less likely to be selected. Partial resolution of medical conditions did not have a significant influence positively or negatively on a patient’s choice of surgery. Each amount of additional weight loss made a surgical profile more likely to be selected, with the loss of 80% of excess body weight as the most significant positive characteristic for the entire study. Each amount of additional cost made a surgical profile less likely to be selected, with an initial out-of-pocket expense of $15 000 as the most significant negative characteristic in the study.

Figure 2. Attribute Variables From Discrete Choice Experiment.

Includes all 815 respondents. A value of greater than 0 indicates that including the associated attribute level makes a surgical profile more likely to be selected. A value less than 0 indicates that the associated attribute level makes a surgical profile less likely to be selected. Data points indicate mean; error bars, SD. HBP indicates high blood pressure; HC, high cholesterol levels.

Table 2. Attribute Variables and Marginal WTP.

Attribute	Coefficient (95% CI)a		Marginal WTP, $b
	Model 1 (Effects Coding)	Model 2 (Alternative Coding)
How treatment works
Reduces calories, increases exercise, behavior modification	0.003 (−0.089 to 0.094)	0.014 (−0.031 to 0.059)	1953
Restricts amount of food, may not decrease hunger	−0.105 (−0.181 to −0.029)c	−0.052 (−0.090 to −0.014)c	NA
Restricts amount of food, decreases hunger	0.079 (−0.010 to 0.168)	0.020 (−0.024 to 0.064)	2138
Restricts amount of food, decreases hunger, reduces absorption	0.023 (−0.035 to 0.082)	0.018 (0.016 to 0.051)	2065
Extensiveness
No recovery, treatment reversible	0.020 (−0.049 to 0.090)	0.006 (−0.033 to 0.044)	4898
1-wk recovery, treatment reversible	−0.071 (−0.134 to −0.007)d	−0.042 (−0.080 to −0.004)d	3492
2-wk recovery, treatment not reversible	0.258 (0.174 to 0.341)e	0.197 (0.148 to 0.246)e	10 534
4-wk recovery, treatment reversible in rare cases	−0.207 (−0.282 to −0.131)e	−0.160 (−0.206 to −0.114)e	NA
Time treatment available, yf
1	−0.086 (−0.148 to −0.025)c	0.029 (0.014 to 0.044)e	852
5	−0.030 (−0.093 to 0.033)
10	−0.049 (−0.113 to 0.015)
20	0.165 (0.097 to 0.234)e
Weight lossg
Lose 40% excess weight, regain 20%, total loss of 20%	−0.473 (−0.563 to −0.384)e	0.185 (0.166 to 0.205)e	5470
Lose 40% excess weight, regain 0%, total loss of 40%	−0.140 (−0.201 to −0.080)e
Lose 80% excess weight, regain 20%, total loss of 60%	0.154 (0.086 to 0.221)e
Lose 80% excess weight, regain 0%, total loss of 80%	0.460 (0.373 to 0.546)e
Resolution of weight-associated medical conditions
Still have diabetes, high cholesterol levels, HBP, or apnea	−0.319 (−0.389 to −0.248)e	−0.207 (−0.254 to −0.159)e	NA
Still have HBP or apnea but no diabetes or high cholesterol levels	−0.026 (−0.101 to 0.048)	−0.038 (−0.073 to −0.002)d	4989
Still have diabetes but no HBP, apnea, or high cholesterol levels	−0.020 (−0.090 to 0.051)	0.016 (−0.026 to 0.057)	6557
No diabetes, high cholesterol levels, HBP, or apnea	0.365 (0.289 to 0.440)e	0.229 (0.177 to 0.280)e	12 843
Risk of complications within 3 y, %h
0	0.172 (0.110 to 0.234)e	−0.072 (−0.092 to −0.052)e	−2133
5	0.081 (0.013 to 0.150)d
20	−0.023 (−0.086 to 0.041)
35	−0.231 (−0.303 to −0.158)e
Adverse effects
None	0.223 (0.143 to 0.303)e	0.169 (0.124 to 0.214)e	7808
Dumping syndrome, diarrhea	0.031 (−0.039 to 0.100)	0.020 (−0.021 to 0.060)	3400
Nausea, vomiting, acid reflux, stomach pains, etc	−0.154 (−0.214 to −0.093)e	−0.096 (−0.130 to −0.061)e	NA
Malabsorption, can result in hair and bone loss	−0.100 (−0.170 to −0.030)c	−0.093 (−0.135 to −0.051)e	71
Diet changes
No restrictions	0.103 (0.039 to 0.167)c	0.071 (0.034 to 0.108)e	4609
Normal foods with small portion, buy vitamins	0.114 (0.044 to 0.183)e	0.081 (0.041 to 0.122)c	4917
Calories restricted, must buy special foods	−0.104 (−0.178 to −0.030)c	−0.067 (−0.113 to −0.021)e	537
Avoid foods high in fat and sugar leves, buy supplements	−0.113 (−0.187 to −0.040)c	−0.085 (−0.127 to −0.043)e	NA
Initial out-of-pocket costs, $i
100	0.309 (0.211 to 0.406)e	−0.034 (−0.042 to −0.025)e	NA
500	0.343 (0.206 to 0.480)e
1000	0.087 (0.001 to 0.173)d
2000	0.107 (0.007 to 0.208)d
3000	0.018 (−0.074 to 0.109)
5000	0.018 (−0.083 to 0.120)
10 000	−0.338 (−0.448 to −0.228)e
15 000	−0.544 (−0.680 to −0.408)e

Abbreviations: HBP, high blood pressure; NA, not applicable; WTP, willingness to pay.

^aValue greater than 0 indicates that including the associated attribute level makes a surgical profile more likely to be selected; less than 0, that the associated attribute level makes a surgical profile less likely to be selected.

^bValue indicates the amount an individual would be willing to pay (if positive) or would have to be paid (if negative) to have the associated attribute level in their surgical profile.

^cP < .01.

^dP < .05.

^eP < .001.

^fCoefficient for continuous model indicates change in preference (or WTP) for every 5 additional years treatment has been available.

^gCoefficient for continuous model indicates change in preference (or WTP) for every 20% of additional total weight loss.

^hCoefficient for continuous model indicates change in preference (or WTP) for every 10% of additional risk of complications within 3 years.

ⁱCoefficient for continuous model indicates change in preference (or WTP) for every $1000 of additional initial out-of-pocket costs.

Other attributes had less influence on respondent choice than the 3 described above. Respondent choice was not greatly influenced by how treatment works, the time a treatment has been available, and the required dietary changes, although certain levels of each of these attributes had positive effects on respondent choice. For example, a profile that had been available for 20 years, had 0% or 5% risk of complications, had no adverse effects, or had no dietary restrictions was more likely to be selected. However, these and many other factors had minimal influence on patient preference when compared with total weight loss, medical conditions, and costs.

Stratified Analyses

Men were more likely to accept a treatment that restricts food and decreases hunger (coefficient value, 0.290 [95% CI, 0.130-0.450] vs 0.030 [95% CI, −0.043 to 0.103]) (eFigure 1 in the Supplement). When compared with older respondents, younger respondents (18-44 years) were more likely to select procedures with greater weight loss (coefficient for loss of 80% of excess weight, 0.543 [95% CI, 0.435-0.651] vs 0.397 [95% CI, 0.315-0.482] for those 45 years or older), procedures with low out-of-pocket costs (coefficient for $100 out-of-pocket costs, 0.346 [95% CI, 0.221-0.470] vs 0.262 [95% CI, 0.174-0.350] for those 45 years or older), and procedures that reduce absorption (coefficient for reducing absorption, 0.122 [95% CI, 0.035-0.208] vs −0.047 [95% CI, −0.107 to 0.013] for those 45 years or older) (eFigure 2 in the Supplement). Although it was not a statistically significant difference, respondents earning $25 000 or less per year were more sensitive to initial out-of-pocket costs than those earning more than $25 000 (coefficient for $100 in out-of-pocket costs, 0.408 [95% CI, 0.217-0.599] vs 0.286 [95% CI, 0.205-0.367]). Otherwise, income level did not significantly affect responses (eFigure 3 in the Supplement).

Marginal Willingness to Pay

Respondents were willing to pay more for procedures that resulted in the greater weight loss ($5470 for every 20% of total weight loss) and that resolved weight-related medical conditions ($12 843; Table 2). They were also willing to pay higher amounts for a nonreversible procedure with a 2-week recovery time ($10 534) and a procedure with no adverse effects ($7808). Respondents showed the least willingness to pay for attributes such as dietary changes, time the treatment has been available, and how the treatment works.

Market Segmentation

Latent class analysis indicated 3 unobserved groups of respondents (Figure 3). The first benefit-focused class was the largest, composed of 312 respondents (38.3%) who were primarily concerned with weight loss and the resolution of weight-associated medical concerns. The second cost-sensitive class was most concerned with out-of-pocket costs and was made up by 256 respondents (31.4%). The decisions of those in the third procedure-focused class of 247 respondents (30.3%) were balanced across all attributes, with slightly greater concern given to how the treatment itself worked. When matched with demographic characteristics, none of the 3 groups were more likely than another to have a particular identity (sex, income, or age).

Figure 3. Latent Class Analysis.

Variables for each class sum to a total of 1. The variable indicates the fraction of influence that an attribute has over selection of a surgical procedure for someone in a given class.

Discussion

This study uses a novel approach, conjoint analysis, to evaluate the attributes of the 4 most common surgical procedures for weight loss in the United States: RYGB, laparoscopic-adjustable gastric banding, SG, and duodenal switch.¹⁷ We find that excess weight loss, initial out-of-pocket costs, and resolution of medical conditions are most important in patient decision making. These preferences differed little by key demographic characteristics, although low-income respondents were more sensitive to costs, younger respondents were more sensitive to excess weight loss, and men were more sensitive to resolution of medical conditions. A marginal willingness-to-pay analysis showed that respondents were willing to spend several thousand dollars for the most important characteristics, such as high excess weight loss and resolution of medical conditions.

One prospective study¹⁸ asked patients the primary reasons for selecting a particular procedure and found different motivations for each; these included safety and least invasiveness for laparoscopic-adjustable gastric banding, the inability to cheat on one’s diet for RYGB, and the amount of weight loss for duodenal switch. A retrospective study¹⁹ found similar variability in reasoning: patients undergoing RYGB cited the evidence base and success rate; those undergoing SG referenced recommendations by physicians; and those undergoing laparoscopic-adjustable gastric banding cited the reversibility and less invasive nature. Yet another study²⁰ found that most patients had made a decision about which procedure to pursue before surgeon consultation and that maximum weight loss was respondents’ primary concern. One study²¹ specifically evaluated the value that patients place on weight loss and found most patients pursued surgery for health reasons and also had unrealistic expectations of weight loss.

Our latent class analysis aligns with the results of previous studies that ask patients why they choose to avoid certain procedures.^19,20,21 These studies have found subgroups of patients who are motivated by 1 or more aspects of a given procedure. Our results show 3 subgroups motivated primarily by cost, by excess weight loss and resolution of medical conditions, or by the reversibility and recovery. As in previous studies,^19,20 these subgroups were not easily divided along demographic lines.

One unique aspect of our study is the importance of initial out-of-pocket costs in patient preferences. Among high-income countries, the influence of cost may be unique to the United States. Most studies from countries with more universal coverage have not taken this influence into account when measuring patients’ choices for bariatric surgery. Findings from the present study suggest that physicians in the United States should strongly take into account the health insurance benefits of their patients.

At present, the 2 most commonly performed operations in the United States are SG and RYGB. Given similar short-term but decreased long-term risks and less technical complexity, SG has become the predominant operation performed in the United States. Furthermore, understanding the concepts of purely restrictive vs a combination of malabsorptive and restrictive regarding how bariatric surgery works has largely been disproven over the last decade.²² Surgeons are often engaged in conversations with their patients about the risks and benefits of each concept, and patients often present with their own preference for an operation. Patient satisfaction after bariatric surgery is of paramount importance because this decision can lead to significant physical and psychosocial changes.¹⁶ The results would not have predicted the rise in popularity of SG compared with RYGB.^4,23 The most likely explanation for this trend is the recommendation of physicians, for which this study does not account.

Limitations

This study has some limitations. The sample was a convenience sample of patients at information sessions in a single state. We do not have information on those who declined to participate in the survey. In addition, respondents completed surveys after participating in information sessions. Therefore, if sessions were biased toward certain risks or benefits, respondent choice could be biased in similar ways. Also, many of the attributes had to be simplified. For example, total weight loss and weight regain were combined. Even after simplifying the list, we had a high number of attributes given that many conjoint analyses will restrict to 5 or fewer attributes. Therefore, the complexity of the choice question may have influenced our results. We were also unable to include many important nonsurgical treatments, such as medications and meal replacements, which would make a worthwhile extension of this study. Finally, the attributes studied herein reflect bariatric surgery at a particular moment in time. The underlying science would no longer be described simply as malabsorption,²² and we may now include endoluminal therapies.

Conclusions

The decision to choose one weight loss method over another is complex and relies on factors related to the procedures, patient preferences, and how physicians frame choices. This study provides information on which characteristics are most important from a patient perspective and that patients demonstrate heterogeneity in preferences. We often know why patients select one procedure over another, but the hypothetical trade-offs in this study give us the relative importance of these attributes. Recognizing that patients will understand these trade-offs differently can improve the discussion between patient and physician on the risks and benefits of various weight loss options.

Supplement.

eTable 1. Description of Attributes and Levels

eTable 2. Attribute Variables for 2 Main Models

eFigure 1. Sensitivity Analysis by Sex

eFigure 2. Sensitivity Analysis by Age

eFigure 3. Sensitivity Analysis by Income