Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2020 Jul 22;155(9):e201985. doi: 10.1001/jamasurg.2020.1985

Association of Roux-en-Y Gastric Bypass With Postoperative Health Care Use and Expenditures in Canada

Jean-Eric Tarride 1,2,3,✉, Aristithes G Doumouras 4,5,6, Dennis Hong 4,5,6, J Michael Paterson 6,7,8, Semra Tibebu 6, Richard Perez 6, Julia Ma 6, Valerie H Taylor 9, Feng Xie 1,2, Vanessa Boudreau 4,5,6, Eleanor Pullenayegum 10,11, David R Urbach 12, Mehran Anvari 4,5,6
PMCID: PMC7376480  PMID: 32697298

This cohort study examines the preoperative and postoperative financial implications of a publicly funded surgical bariatric procedure.

Key Points

Question

Is a Roux-en-Y gastric bypass procedure associated with higher health care spending 5 years after the procedure compared with a matched control group in Ontario, Canada?

Findings

In this population-based cohort study of 1587 patients who underwent a Roux-en-Y gastric bypass and 1587 control individuals, the net health care expenditures associated with the procedure were CAD $10 831 (2017 Canadian dollars) (US $8341) over 5 years, excluding the costs associated with the date of the procedure. Health care expenditures were statistically significantly higher during the 3 years after the procedure but were similar thereafter to spending in the control group.

Meaning

These findings suggest that in a setting with a public health care system, Roux-en-Y gastric bypass is associated with increased health care costs in the short term.

Abstract

Importance

Results of previous studies are mixed regarding the economic implications of a Roux-en-Y gastric bypass (RYGB).

Objective

To assess the 5-year incremental health care use and expenditures after RYGB.

Design, Setting, and Participants

This population-based cohort study conducted in Ontario, Canada, used a difference-in-differences approach to compare health care use and expenditures between patients who underwent a publicly funded RYGB from March 1, 2010, to March 31, 2013, and propensity score–matched control individuals who did not undergo a surgical bariatric procedure. The study period allowed for a minimum 60 months of follow-up because, at that time, the most recent date for which administrative data on health care and expenditures were available was March 31, 2018. Data sources included the Ontario Bariatric Registry linked to several Ontario health administrative databases and the Electronic Medical Record Administrative Data Linked Database. Health care use and expenditures data for 5 years before and 5 years after the index date (procedure date for RYGB group; random date for controls) were analyzed. Data analyses were performed March 12, 2019, to March 10, 2020.

Intervention

RYGB procedure.

Main Outcomes and Measures

The primary outcome was total health care expenditures.

Results

The final propensity score–matched cohorts comprised 1587 individuals in the RYGB group (mean [SD] age, 47 [10.2] years) and 1587 controls (mean [SD] age, 47 [12.2] years); each group had 1228 women (77.4%) and a mean body mass index (calculated as weight in kilograms divided by height in meters squared) of 46. Mean total health care expenditures (2017 Canadian dollars) per patient in the RYGB group increased from CAD $15 594 (95% CI, CAD $14 743 to CAD $16 614) (US $12 008 [95% CI, US $11 353 to US $12 794]) in the 5 years before the procedure to CAD $30 389 (95% CI, CAD $28 789 to CAD $32 232) (US $23 401 [95% CI, US $22 169 to US $24 821]) over the 5 years after the procedure, a difference of CAD $14 795 (95% CI, CAD $13 172 to CAD $16 480) (US $11 393 [95% CI, US $10 143 to US $12 691]). For the control group, mean total health care expenditures per individual increased from CAD $16 109 (95% CI, CAD $14 727 to CAD $17 591) (US $12 405 [95% CI, US $11 341 to US $13 546]) 5 years before the index date to CAD $20 073 (95% CI, CAD $18 147 to CAD $22 169) (US $15 457 [95% CI, US $13 974 to US $17 071]) 5 years after the date, a difference of CAD $3964 (95% CI, CAD $2250 to CAD $5875) (US $3053 [95% CI, US $1733 to US $4524]). Overall, the difference-in-differences estimate of the net cost of RYGB was CAD $10 831 (95% CI, CAD $8252 to CAD $13 283) (US $8341 [95% CI, $6355 to $10 229]) over the 5-year period. This amount excluded the mean (SD) cost associated with the index date: CAD $6501 (CAD $1087) (US $5006 [US $837]) for the RYGB cohort and CAD $9 (CAD $72) (US $7 [US $55]) for the controls. The cost differential was primarily associated with increased hospitalizations in the first months immediately after RYGB. Expenditures leveled off in year 3 after the index date; differences in total expenditures between the RYGB and control cohorts were not statistically significantly different in years 4 and 5.

Conclusions and Relevance

Health care expenditures in the 3 years after publicly funded RYGB were higher in patients who underwent the procedure than in control individuals, but the costs were similar thereafter. This finding suggests the need to decrease hospital and emergency department readmissions after surgical bariatric procedures because such use is associated with increased spending.

Introduction

The sharp increase in the prevalence of morbid obesity is associated with increased requests for surgical bariatric procedures.1 Compared with nonsurgical interventions, surgical bariatric procedures can have dramatic outcomes, such as reversal of or substantial reduction in comorbidities (eg, type 2 diabetes) and mortality.2,3,4,5,6,7,8,9,10,11 Despite these health outcomes, results from previous research have been mixed regarding the economic implications of surgical bariatric procedures. Some studies have reported an increase in health care expenditures after surgical bariatric interventions,12,13,14,15,16,17,18,19 whereas other studies have shown postoperative savings in the general population13,20,21,22,23,24,25,26,27,28,29,30 and among patients with diabetes.31,32,33

With 14.3 million residents in 2018, the Canadian province of Ontario represents approximately 40% of the national population.34 Ontario residents with a body mass index (BMI [calculated as weight in kilograms divided by height in meters squared]) of 40 or higher or a BMI of 35 to 39 with obesity-related comorbid conditions are eligible for publicly funded surgical bariatric procedures (ie, Roux-en-Y gastric bypass [RYGB] and sleeve gastrectomy); adjustable gastric banding is not currently reimbursed by the Ontario Ministry of Health. Overall, the Ontario Ministry of Health pays almost 100% of hospital care, emergency department (ED) visits, outpatient clinics, and physician visits for all residents of Ontario and 100% of prescription drugs for seniors and social assistance recipients. Public subsidy levels are lower for other services, such as home care.

To better inform physicians, patients, and payers about the value of publicly funded RYGB, we conducted a study in Ontario, Canada, to identify the health care use and expenditures of individuals 5 years after undergoing an RYGB compared with a matched control group (who did not undergo surgical bariatric interventions). We hypothesized that health care expenditures between patients with an RYGB and control individuals would not differ in year 5.

Methods

This cohort study was approved by the Research Ethics Board at Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada. Informed consent was not obtained because the study used data maintained by ICES, an independent, nonprofit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation. In analyzing and reporting these data, we followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.35

Data Sources

Since 2010, the Ontario Bariatric Registry has been collecting data on all surgical patients eligible for publicly funded surgical bariatric procedures.36,37,38,39,40,41 For the purpose of the present study, we transferred and linked the Ontario Bariatric Registry to several health administrative databases (see eTable 1 in the Supplement) housed at ICES. To identify nonsurgical control individuals, we used the Electronic Medical Record Administrative Data Linked Database (EMRALD),42 a primary care database of more than 600 000 residents of Ontario that contains BMI information.43 These data sets were linked using unique encoded identifiers and analyzed at ICES.

Study Populations

Using the Ontario Bariatric Registry, we identified all individuals who underwent publicly funded RYGB from March 1, 2010, to March 31, 2013. This period allowed for a minimum of 5 years (60 months) of follow-up because, at the time of the study, the most recent date for which administrative data on health care use and expenditures were available was March 31, 2018. To minimize the risk of choosing control individuals with a single BMI measurement from a previous event (eg, weight and height collected after a hospitalization), we used the EMRALD database to select from a pool of individuals with multiple BMI measurements taken between January 1, 2010, and December 31, 2014. Next, we chose a random BMI measurement and assigned an index date that was approximately 365 days from that BMI measurement.

Outcomes

The primary study outcome was total health care expenditure, which was calculated using the ICES costing algorithms.44 The secondary outcome included health care use, which was defined as the number of hospitalizations (overall and by admission type [elective and nonelective]), ED visits, same-day surgical procedures, hospital outpatient clinic visits, general practitioner and specialist visits, mental health admissions, home care services used, and dialysis and cancer clinic visits. Prescription medications and associated costs were not included because only publicly reimbursed medications for social assistance recipients and those aged 65 years or older were available in the administrative data. Health insurance status and date of death were obtained from the Ontario Health Insurance Plan Registered Persons Database. For the present analyses, we included 5-year mortality after the index date as a secondary outcome.

Statistical Analysis

We used a propensity score45,46 to identify a comparison cohort of control individuals who did not receive bariatric surgical interventions but were otherwise similar in characteristics to patients who underwent an RYGB. Similar to previous studies,47,48 we used logistic regression to compute the probability of receipt of surgical bariatric procedures among all surgical and potential control individuals as a function of age, sex, BMI, index date (date of RYGB; random date for control group), geographical location (ie, 14 local health integration networks), census neighborhood income quintile, Ontario Marginalization Index, number of major Aggregated Diagnostic Groups (derived from the Johns Hopkins ACG System, version 10), medical conditions derived from validated ICES administrative data algorithms (chronic kidney disease, coronary artery disease, diabetes, hypertension, hypercholesterolemia, and mood and anxiety disorders), total health care expenditures in the 5 years preceding the index date, and the number of days in hospital and number of ED visits in the 365 days preceding the index date. Each patient with an RYGB was matched to a nonsurgical patient using greedy nearest-neighbor matching, which matches individuals on the basis of the logit of their propensity score and surgical status using a caliper width of 0.2 of the SD. Covariate balance between the RYGB group and control group was evaluated using standardized differences.49 In general, standardized differences greater than 0.1 are considered meaningful. Matched and unmatched individuals within each cohort were also compared, using t tests and χ2 tests for unpaired data. Two-sided P < .05 were considered statistically significant.

The core statistical analysis was based on the difference-in-differences approach, which is often used to test for intervention effects.50,51,52,53,54,55 Under this approach, differences in outcomes before and after the index date are calculated for the RYGB group (D1) before being calculated for the control group (D2). In the absence of an association between surgical bariatric procedures and outcomes, the difference in differences (D1 – D2) is equal to 0. This method allows the identification of the change in outcomes that is associated with the intervention and is beyond secular trends. To facilitate interpretation, we presented the results of the difference-in-differences approach by comparing the outcomes 5 years before and 5 years after the index date. Estimates from the difference-in-differences approach were bootstrapped56 to generate CIs around the estimates and to preserve the matching and the repeated measures structure of the data. To provide additional insights, we used the generalized linear model with γ distributions and log-link functions57 to compare the costs between the matched RYGB and control cohorts for each year separately after the index date. Because of the study design, the costs associated with the index date were not included in the difference-in-differences estimates but were documented for the 2 groups.

All base case analyses were calculated on an intention-to-treat basis and included patients who died or those who moved out of Ontario and thus lost their Ontario health insurance status. Sensitivity analyses were conducted with matched pairs in which both the RYGB and control groups had complete data at the end of the 60-month follow-up period. All costs were expressed in 2017 Canadian dollars. SAS Enterprise Guide, version 7.1 (SAS Institute Inc), was used for the analyses. Data analyses were performed from March 12, 2019, to March 10, 2020.

Results

The initial (before matching) eligible populations comprised 2980 Ontario residents who underwent publicly funded RYGB and 6037 potential control individuals with BMI data who did not undergo surgical bariatric interventions. The RYGB group received 93% of all publicly funded surgical bariatric procedures conducted in Ontario before March 2013. Compared with the RYGB cohort, the potential control cohort was older (aged ≥65 years: 2% vs 27%; P < .001), had a lower proportion of women (84% vs 60%; P < .001), and had a lower mean BMI (48 vs 42; P < .001). eFigure 1 in the Supplement presents the study flow diagram, and eTable 2 (RYGB cohort) and eTable 3 (control cohort) in the Supplement present the full characteristics of the matched and unmatched groups.

After propensity score matching, the final study sample included 1587 patients in the RYGB group and 1587 individuals in the control group. As shown in Table 1, the matched groups were well balanced. The mean (SD) age was 47 (10.2) years for the RYGB cohort and 47 (12.2) years for the control cohort, and both groups comprised 1228 women (77.4%) and 359 men (22.6%). The mean BMI was 46, and approximately one-third of individuals had diabetes (545 [34.3%] for the RYGB group and 501 [31.6%] for the control group) (Table 1).

Table 1. Baseline Characteristics of the Matched Cohortsa.

Characteristic No. (%) Standardized difference
RYGB group (n = 1587) Control group (n = 1587)
Age, mean (SD), y 47 (10.2) 47 (12.2) 0
<45 634 (39.9) 637 (40.1) 0
45-54 536 (33.8) 528 (33.3) 0.01
55-64 366 (23.1) 359 (22.6) 0.01
≥65 51 (3.2) 63 (4.0) 0.04
Sex
Female 1228 (77.4) 1228 (77.4) 0
Male 359 (22.6) 359 (22.6) 0
BMI, mean (SD) 46.55 (6.22) 46.22 (13.60) 0.03
Income quintile
Lowest 325 (20.5) 359 (22.6) 0.05
2 343 (21.6) 346 (21.8) 0
3 340 (21.4) 317 (20.0) 0.04
4 334 (21.0) 337 (21.2) 0
Highest 243 (15.3) 227 (14.3) 0.03
Missing 2 (0.1) 1 (0.1) 0.02
Marginalization index summary score, mean (SD) 2.97 (0.76) 3.02 (0.78) 0.07
Total No. of major ADG, mean (SD) 5.3 (2.5) 5.4 (3.0) 0.02
Select medical conditions in preceding 5 y
CKD/ESKD 41 (2.6) 32 (2.0) 0.04
CAD/PCI/CABG 222 (14.0) 216 (13.6) 0.01
Diabetes 545 (34.3) 501 (31.6) 0.06
Hypertension 659 (41.5) 638 (40.2) 0.03
Hypercholesterolemia 285 (18.0) 253 (15.9) 0.05
Mood and anxiety disorders 24 (1.5) 25 (1.6) 0.01
Total health care expenditure in preceding 5 y, mean (SD), CAD$b 15 594 (19 820) 16 109 (29 529) 0.02
Open in a new tab

Abbreviations: ADG, aggregated diagnosis group; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CABG, coronary artery bypass graft; CAD, coronary artery disease; CKD, chronic kidney disease; ESKD, end-stage kidney disease; PCI, percutaneous coronary intervention; RYGB, Roux-en-Y gastric bypass.

a

In addition, the cohorts were matched according to 14 local health integration networks of residence.

b

To convert 2017 Canadian dollars to 2017 US dollars, divide by 1.2986.

Mean (SD) total health care expenditures in the 5-year period before the index date were CAD $15 594 (CAD $19 820) (2017 Canadian dollars) (US $12 008 [US $15 263]) per individual for the RYGB group and CAD $16 109 (CAD $29 529) (US $12 405 [US $22 739]) per individual for the control group (Table 1), which were not statistically significantly different between the groups (P = .56). Approximately 97% of the study population had 60 months of follow-up data: 1541 of 1587 patients for the RYGB group [97.1%] and 1531 of 1587 individuals for the control group [96.5%]). Over the 5-year period, 21 patients in the RYGB cohort and 11 in the control cohort lost their health insurance status, whereas 25 patients in the RYGB group and 45 in the control group died. The cumulative mortality at 5 years was 1.6% (n = 25) for the RYGB group and 2.8% (n = 44) for the nonsurgical group (P = .02).

Health Care Use and Expenditures 5 Years Before and After Index Date

Although health care use and expenditures were comparable between the matched cohorts before the index date, they differed afterward. Trends over time are presented for the number of hospitalizations in Figure 1A, ED visits in Figure 1B, specialist visits in Figure 2A, and total health care expenditures in Figure 2B for both cohorts. As shown in these graphs, a sharp increase was observed in the mean number of hospitalizations (from 0.01 to 0.10), ED visits (from 0.13 to 0.43), specialist visits (from 5.22 to 6.51), and total spending (from CAD $1667 to CAD $9207 [US $1284 to US $7090]) in the 3 months before and after an RYGB. The mean number of specialist visits in the RYGB group also increased from the 10- to 12-month period before the index date (1.83 visits) to the 3 months before the index date (5.22 visits).

Figure 1. Trends for Hospitalizations and Emergency Department (ED) Visits.

Figure 1.

Open in a new tab

Vertical line with 0 indicates the index date, with the numbers on its right marking the period after the index date and the (negative) numbers on its left marking the period before the index date. Mean numbers of hospitalizations (A) and ED visits (B) were based on a total number of 1587 individuals before the index date (time 0) for both Roux-en-Y gastric bypass (RYGB) and control groups. For the RYGB group, the numbers of patients were 1587 for year 1, 1579 for year 2, 1571 for year 3, 1564 for year 4, and 1552 for year 5, with 1541 having 60 months of data. For the control group, the numbers of individuals were 1587 for year 1, 1584 for year 2, 1575 for year 3, 1566 for year 4, and 1556 for year 5, with 1531 having 60 months of data.

Figure 2. Trends for Specialist Visits and Total Health Care Expenditures.

Figure 2.

Open in a new tab

Vertical line with 0 indicates the index date, with the numbers on its right marking the period after the index date and the (negative) numbers on its left marking the period before the index date. Mean numbers of specialist visits (A) and total health care expenditures (B) were based on a total number of 1587 individuals before the index date (time 0) for both Roux-en-Y gastric bypass (RYGB) and control groups. For the RYGB group, the numbers of patients were 1587 for year 1, 1579 for year 2, 1571 for year 3, 1564 for year 4, and 1552 for year 5, with 1541 having 60 months of data. For the control group, the numbers of individuals were 1587 for year 1, 1584 for year 2, 1575 for year 3, 1566 for year 4, and 1556 for year 5, with 1531 having 60 months of data.

Table 2 presents health care use and expenditures for each cohort 5 years before and 5 years after the index date and the resulting difference-in-differences approach estimates. Largely associated with the increase in hospitalizations in the first months after the surgical procedure, the mean total health care expenditures per patient in the RYGB group increased from CAD $15 594 (95% CI, CAD $14 743 to CAD $16 614) (US $12 008 [95% CI, US $11 353 to US $12 794]) in the 5-year period before the procedure to CAD $30 389 (95% CI, CAD $28 789 to CAD $32 232) (US $23 401 [95% CI, US $22 169 to US $24 821]) over the 5-year period after, for a difference of CAD $14 795 (95% CI, CAD $13 172 to CAD $16 480) (US $11 393 [95% CI, US $10 143 to US $12 691]). In contrast, for the control group, the mean total health care expenditures per individual increased from CAD $16 109 (95% CI, CAD $14 727 to CAD $17 591) (US $12 405 [95% CI, US $11 341 to US $13 546]) 5 years before the index date to CAD $20 073 (95% CI, CAD $18 147 to CAD $22 169) (US $15 457 [95% CI, US $13 974 to US $17 071]) 5 years afterward, for a difference of CAD $3964 (95% CI, CAD $2250 to CAD $5875) (US $3053 [95% CI, US $1733 to US $4524]). Overall, the difference-in-differences estimate of the net cost of an RYGB for total health care expenditures was CAD $10 831 (95% CI, CAD $8252 to CAD $13 283) (US $8341 [95% CI, $6355 to $10 229]) over the 5-year period (Table 2). This estimate did not include the mean (SD) costs per individual associated with the index date, which were CAD $6501 (CAD $1087) (US $5006 [US $837]) for the RYGB cohort and CAD $9 (CAD $72) (US $7 [US $55]) for the control cohort. In terms of the net implication of an RYGB for 5-year health care use, the results of the difference-in-differences calculations indicate that the RYGB group consumed fewer health care resources than the control group in terms of general practitioner visits (difference-in-differences: −2.55; 95% CI, −4.36 to −0.68) and home care services (difference-in-differences: −6.49; 95% CI, −11.39 to −2.03) (Table 2).

Table 2. Difference-in-Differences Estimate in Health Care Expenditures and Usea,b.

Variable RYGB group Control group Difference-in-differences estimate (95% CI)c
5 y Before index date, (95% CI) 5 y After index date, (95% CI) 5 y Before index date, (95% CI) 5 y After index date, (95% CI)
Health care expenditures, CAD$d 15 594 (14 743 to 16 614) 30 389 (28 789 to 32 232) 16 109 (14 727 to 17 591) 20 073 (18 147 to 22 169) 10 831 (8252 to 13 283)
No. of ED visits 3.36 (3.10 to 3.63) 4.44 (4.11 to 4.81) 3.38 (3.14 to 3.63) 3.39 (3.13 to 3.67) 1.07 (0.69 to 1.47)
No. of hospitalizations 0.41 (0.37 to 0.46) 0.97 (0.90 to 1.05) 0.39 (0.34 to 0.42) 0.52 (0.47 to 0.59) 0.42 (0.32 to 0.53)
Elective 0.20 (0.17 to 0.22) 0.34 (0.31 to 0.38) 0.17 (0.15 to 0.20) 0.18 (0.16 to 0.21) 0.14 (0.09 to 0.19)
Nonelective 0.21 (0.18 to 0.26) 0.63 (0.57 to 0.70) 0.21 (0.18 to 0.24) 0.34 (0.29 to 0.40) 0.29 (0.20 to 0.37)
No. of visits
SDS 1.04 (0.98 to 1.10) 1.43 (1.31 to 1.57) 0.96 (0.88 to 1.06) 0.97 (0.90 to 1.05) 0.38 (0.22 to 0.53)
GP 36.70 (35.49 to 38.03) 33.53 (32.26 to 34.87) 36.62 (35.03 to 38.22) 35.99 (34.20 to 37.82) −2.55 (−4.36 to −0.68)
Specialist 35.52 (34.19 to 36.94) 45.09 (42.94 to 47.42) 31.75 (29.73 to 33.81) 39.06 (36.70 to 41.45) 2.25 (−0.96 to 5.23)
Cancer clinic 0.18 (0.08 to 0.30) 0.36 (0.19 to 0.56) 0.43 (0.28 to 0.59) 0.61 (0.42 to 0.84) −0.01 (−0.32 to 0.31)
Dialysis clinic 0.37 (0.00 to 1.04) 0.59 (0.08 to 1.32) 0.40 (0.00 to 0.90) 0.91 (0.16 to 2.01) −0.29 (−1.14 to 0.45)
Outpatient 8.96 (8.51 to 9.42) 7.69 (7.16 to 8.25) 8.00 (7.32 to 8.70) 7.03 (6.44 to 7.65) −0.30 (−1.08 to 0.51)
No. of home care services 4.73 (3.13 to 6.69) 7.71 (5.86 to 10.04) 9.41 (6.34 to 13.34) 18.89 (13.59 to 25.05) −6.49 (−11.39 to −2.03)
No. of mental health admissions 0.03 (0.02 to 0.04) 0.07 (0.04 to 0.11) 0.09 (0.06 to 0.12) 0.08 (0.05 to 0.11) 0.05 (0.01 to 0.10)
Open in a new tab

Abbreviations: ED, emergency department; GP, general practitioner; RYGB, Roux-en-Y gastric bypass; SDS, same-day surgery.

a

Numbers represent bootstrapped means (CIs) of health care expenditures and use per individual.

b

The mean (SD) of the costs per individual on the index date were CAD $6501 ($1087) (US $5006 [US $837]) for the RYGB group and CAD $9 (CAD $72) (US $7 [US $55]) for the control group (not included in calculations).

c

Difference-in-differences estimate was based on 1541 patients in the RYGB group and 1531 individuals in the control group with 60 months of data (of 1587 people before the index date [t = 0] for both groups). The estimate was calculated as the difference in health care expenditures and use before and after the index date for the RYGB group minus the difference in health care expenditures and use before and after the index date for the control group.

d

To convert 2017 Canadian dollars to 2017 US dollars, divide by 1.2986.

Over time, total health care expenditures decreased for the RYGB group and increased slightly for the control group, leveling off in year 3 after the index date (Figure 3). The differences in mean (SD) total health care expenditures between the RYGB and control cohorts were not statistically significantly different in year 4 (CAD $4188 [CAD $9016] vs CAD $4017 [CAD $12 661] [US $3225 (US $6943) vs US $3093 (US $9750)]; P = .35) and in year 5 (CAD $4100 [CAD $10 078] vs CAD $4023 [CAD $12 923] [US $3157 (US $7761) vs US $3098 (US $9951)]; P = .79) after the index date. eTable 4 (RYGB cohort) and eTable 5 (control cohort) in the Supplement present the total health care expenditures and individual cost components for each year of the analysis.

Figure 3. Total Health Care Expenditures for Matched Patients With Roux-en-Y Gastric Bypass (RYGB) and Control Individuals per Year of Analysis.

Figure 3.

Open in a new tab

The whiskers indicate the mean health care expenditures and CIs at each period for both RYGB and control groups. For year 1, mean (SD) numbers do not include the costs associated with the index date, which were CAD $6501 (CAD $1087) (US $5006 [US $837]) for the patients who underwent RYGB and CAD $9 (CAD $72) (US $7 [US $55]) for the control individuals. Data were based on 1587 individuals before the index date (time 0) for both groups. For the RYGB group, the numbers of patients were 1587 for year 1, 1579 for year 2, 1571 for year 3, 1564 for year 4, and 1552 for year 5, with 1541 having 60 months of data. For the control group, the numbers of individuals were 1587 for year 1, 1584 for year 2, 1575 for year 3, 1566 for year 4, and 1556 for year 5, with 1531 having 60 months of data. To convert mean health care expenditures from 2017 Canadian dollars to 2017 US dollars, divide by 1.2986.

Sensitivity Analysis

By including only the individuals with 60-month follow-up data (eg, excluding patients who died or lost their health insurance status), the 5-year difference-in-differences cost estimate in the base case slightly increased from CAD $10 831 (95% CI, CAD $8252 to CAD $13 283) (US $8341 [95% CI, $6355 to $10 229]) (Table 2) to CAD $10 974 (95% CI, CAD $8410 to CAD $13 267) (US $8451 [95% CI, US $6476 to US $10 216]). eTable 6 in the Supplement presents the results of the sensitivity analysis.

Discussion

To our knowledge, this cohort study is the first contemporary large Canadian population–based analysis of health care use and expenditures in the 5 years before and 5 years after an RYGB. Using a matched control group, we found that the net cost associated with an RYGB was CAD $10 831 (US $8341) per individual over 5 years, when excluding the cost associated with the day of the surgical procedure (ie, CAD $6501 [US $5006]). The additional expenditure was mostly associated with an increase in hospitalizations in the first months after the surgical bariatric intervention. Total health care expenditures decreased over time in the RYGB cohort and were not statistically significantly different between the matched groups in years 4 and 5 after the index date. In parallel, we observed a statistically significant decrease in mortality in favor of the RYGB group.

This study differs from previous evaluations in several ways. Compared with studies with a pre–post design but without referencing a nonsurgical control group12,13,14,21,23,25,26,27,28,33 or studies with matched controlled groups but without controlling for past health care expenditures,15,16,31,32,58,59 the present study used a difference-in-differences approach with 5 years of data before and after an RYGB and with a propensity score–matched controlled group. The difference-in-differences approach requires that the assumption of parallel trends in outcomes before the surgical intervention be satisfied, which, in this case, was satisfied by matching on previous health care expenditures. Although controlling for past health care expenditures could introduce bias associated with evaluation costs in the year preceding the surgical bariatric procedure, we believe that the implication for the study results is likely to be small for 2 reasons. First, we controlled for total health care expenditures over 5 years rather than just in the year immediately preceding the RYGB. As a result, the data showed that the number of specialist visits increased in the RYGB group 1 year before the procedure (Figure 2A), a finding that is consistent with real practice. Second, we observed no important differences between the matched and unmatched cohorts in terms of 5-year health care expenditures before the index date, suggesting that the bias from matching on past health care expenditures should be small. Compared with registry studies with a high attrition rate over time, this study had a lost to follow-up rate (attributed to discontinued provincial health insurance coverage) of less than 1% of the population because we linked the Ontario Bariatric Registry to the administrative databases in the province. As such, we had access to a rich population-based longitudinal data set with 5 years of follow-up data, allowing us to better understand the implication of a publicly funded RYGB for health care use and expenditures over time. Furthermore, we provided a detailed description of the costs incurred by the study populations, including some costs (eg, mental health admissions, home care services, dialysis services, cancer clinic visits, and laboratory tests) that are rarely described in the literature on the economics of surgical bariatric procedures.

Although comparing the present study with research conducted in different jurisdictions or with different study designs (eg, pre–post vs controlled) is difficult, we believe that the results are aligned with those of a few studies that used 5 years or more of follow-up data that have shown that health care expenditures increased after the first years of surgical bariatric intervention and were comparable to the control group thereafter.15,16,19,60 However, most of these studies were conducted in the US16,19 or in Brazil,60 and it is not clear how their results would apply to publicly funded health care such as the system in Canada. Although conducted in a nation with a public health care system, the Swedish Obese Subjects study15 may not reflect the current bariatric landscape. As such, the present study confirms the findings in the Swedish Obese Subjects study and in the US studies as well as highlights some of the limitations associated with studies that use a shorter time horizon (eg, 3 years) in evaluating the economic implications of surgical bariatric procedures.

Consistent with findings in other studies,61 these results highlight the need for strategies to decrease hospital and ED readmissions after surgical interventions. Such use is associated with the incremental increase in health care expenditures after these procedures.

Limitations

This study has some limitations. First, although we had access to all publicly funded RYGB procedures and comprehensive administrative databases showing all publicly funded health care expenditures in Ontario, it is likely that we underestimated the financial advantages of these surgical bariatric procedures because we did not account for savings in prescribed medications after the procedure. For example, a recent US study reported that 4-year postoperative pharmacy costs were significantly lower among 2700 patients who underwent the surgical bariatric intervention compared with 2700 matched control individuals (US $8411 vs $9900; P < .001).62 In Ontario, prescribed medications are reimbursed by the Ontario Ministry of Health and Long-Term Care for older adults and social assistance recipients only, whereas other Canadian citizens are covered by private insurance or pay out-of-pocket. We did not have access to privately funded drug expenditures data. Second, nearly 50% of the RYGB population was lost during the matching process because the database for the control individuals had limited BMI data and because of the differences in patient characteristics between the pool of patients who underwent an RYGB and the pool of control individuals. Twenty-seven percent of the control pool comprised those aged 65 years or older, whereas only 2% of the RYGB pool was composed of older adults. This age difference and the lack of access to population-based BMI data reduced the number of individuals available for matching. However, no differences in health care costs were found between the matched and unmatched patients with an RYGB both before and after the procedure (eFigure 2 in the Supplement), indicating that the results are likely to be representative of the health care expenditures associated with the population undergoing an RYGB in Ontario. Third, although we included many potential confounders that yielded a good match, the potential for residual confounding always exists in any observational study. Because we evaluated only RYGB, the results may not be generalizable to settings in which gastric banding or sleeve gastrectomy is prevalent. We did not document the reasons for the postoperative admissions, and although some of these admissions may have been associated with complications after an RYGB, others may have been procedures (eg, knee replacement) deferred because of the extra body weight prior to the surgical procedure. Fourth, we were unable to incorporate quality of life or indirect costs in the calculations to reflect a broader societal perspective. Such factors are important to consider in addition to saving lives and decreasing morbidity because surgical bariatric procedures have been associated with improved quality of life 63 and social transfers.64

Conclusions

This population-based cohort study in Ontario, Canada, found that the costs of publicly funded RYGB appeared to be higher in the first 3 years but similar thereafter to the costs of nonsurgical bariatric interventions. In parallel, mortality decreased in the RYGB group, highlighting some of the societal implications of surgical bariatric procedures. Results of this study suggest the need for strategies to decrease hospital and ED readmissions after surgical bariatric interventions because such use is associated with increased health care expenditures.

Supplement.

eTable 1. Administrative Data Sources

eFigure 1. Study Flow Diagram

eTable 2. Comparisons of Unmatched and Matched RYGB Cohorts

eTable 3. Comparisons of Unmatched and Matched Controls

eTable 4. Total Healthcare Expenditures and Individual Cost Components for Matched RYGB Patients, Per Year of Analysis

eTable 5. Total Healthcare Expenditures and Individual Cost Components for Matched Controls, Per Year of Analysis

eTable 6. Difference-in-Differences (DID) in Healthcare Expenditures and Resource Utilization for Patients With Complete Data at Month 60

eFigure 2. Costs Before and After Surgery for Unmatched and Matched RYGB Patients

Click here for additional data file. (334.6KB, pdf)

References

  • 1.Baker MT. The history and evolution of bariatric surgical procedures. Surg Clin North Am. 2011;91(6):1181-1201, viii. doi: 10.1016/j.suc.2011.08.002 [DOI] [PubMed] [Google Scholar]
  • 2.Puzziferri N, Roshek TB III, Mayo HG, Gallagher R, Belle SH, Livingston EH. Long-term follow-up after bariatric surgery: a systematic review. JAMA. 2014;312(9):934-942. doi: 10.1001/jama.2014.10706 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Colquitt JL, Pickett K, Loveman E, Frampton GK. Surgery for weight loss in adults. Cochrane Database Syst Rev. 2014;8(8):CD003641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Courcoulas AP, Yanovski SZ, Bonds D, et al. Long-term outcomes of bariatric surgery: a National Institutes of Health symposium. JAMA Surg. 2014;149(12):1323-1329. doi: 10.1001/jamasurg.2014.2440 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Padwal R, Klarenbach S, Wiebe N, et al. Bariatric surgery: a systematic review of the clinical and economic evidence. J Gen Intern Med. 2011;26(10):1183-1194. doi: 10.1007/s11606-011-1721-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Chang SH, Stoll CR, Song J, Varela JE, Eagon CJ, Colditz GA. The effectiveness and risks of bariatric surgery: an updated systematic review and meta-analysis, 2003-2012. JAMA Surg. 2014;149(3):275-287. doi: 10.1001/jamasurg.2013.3654 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.O’Brien PE, Hindle A, Brennan L, et al. Long-term outcomes after bariatric surgery: a systematic review and meta-analysis of weight loss at 10 or more years for all bariatric procedures and a single-centre review of 20-year outcomes after adjustable gastric banding. Obes Surg. 2019;29(1):3-14. doi: 10.1007/s11695-018-3525-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Sjöström L. Review of the key results from the Swedish Obese Subjects (SOS) trial—a prospective controlled intervention study of bariatric surgery. J Intern Med. 2013;273(3):219-234. doi: 10.1111/joim.12012 [DOI] [PubMed] [Google Scholar]
  • 9.Sjöström L, Gummesson A, Sjöström CD, et al. ; Swedish Obese Subjects Study . Effects of bariatric surgery on cancer incidence in obese patients in Sweden (Swedish Obese Subjects Study): a prospective, controlled intervention trial. Lancet Oncol. 2009;10(7):653-662. doi: 10.1016/S1470-2045(09)70159-7 [DOI] [PubMed] [Google Scholar]
  • 10.Sjöström L, Narbro K, Sjöström CD, et al. ; Swedish Obese Subjects Study . Effects of bariatric surgery on mortality in Swedish obese subjects. N Engl J Med. 2007;357(8):741-752. doi: 10.1056/NEJMoa066254 [DOI] [PubMed] [Google Scholar]
  • 11.Sjöström L, Peltonen M, Jacobson P, et al. Bariatric surgery and long-term cardiovascular events. JAMA. 2012;307(1):56-65. doi: 10.1001/jama.2011.1914 [DOI] [PubMed] [Google Scholar]
  • 12.Bleich SN, Chang HY, Lau B, et al. Impact of bariatric surgery on health care utilization and costs among patients with diabetes. Med Care. 2012;50(1):58-65. doi: 10.1097/MLR.0b013e3182290349 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Czernichow S, Moszkowicz D, Szwarcensztein K, et al. Impact of bariatric surgery on the medical management and costs of obese patients in France: an analysis of a national representative claims database. Obes Surg. 2015;25(6):986-996. doi: 10.1007/s11695-014-1488-3 [DOI] [PubMed] [Google Scholar]
  • 14.Maciejewski ML, Smith VA, Livingston EH, et al. Health care utilization and expenditure changes associated with bariatric surgery. Med Care. 2010;48(11):989-998. doi: 10.1097/MLR.0b013e3181ef9cf7 [DOI] [PubMed] [Google Scholar]
  • 15.Neovius M, Narbro K, Keating C, et al. Health care use during 20 years following bariatric surgery. JAMA. 2012;308(11):1132-1141. doi: 10.1001/2012.jama.11792 [DOI] [PubMed] [Google Scholar]
  • 16.Weiner JP, Goodwin SM, Chang HY, et al. Impact of bariatric surgery on health care costs of obese persons: a 6-year follow-up of surgical and comparison cohorts using health plan data. JAMA Surg. 2013;148(6):555-562. doi: 10.1001/jamasurg.2013.1504 [DOI] [PubMed] [Google Scholar]
  • 17.Bhatti JA, Nathens AB, Thiruchelvam D, Redelmeier DA. Weight loss surgery and subsequent emergency care use: a population-based cohort study. Am J Emerg Med. 2016;34(5):861-865. doi: 10.1016/j.ajem.2016.02.007 [DOI] [PubMed] [Google Scholar]
  • 18.Canadian Institute for Health Information Bariatric surgery in Canada. Published May 2014. Accessed June 17, 2020. https://secure.cihi.ca/free_products/Bariatric_Surgery_in_Canada_EN.pdf
  • 19.Smith VA, Arterburn DE, Berkowitz TSZ, et al. Association between bariatric surgery and long-term health care expenditures among veterans with severe obesity. JAMA Surg. 2019;154(12):e193732. doi: 10.1001/jamasurg.2019.3732 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Cremieux PY, Buchwald H, Shikora SA, Ghosh A, Yang HE, Buessing M. A study on the economic impact of bariatric surgery. Am J Manag Care. 2008;14(9):589-596. [PubMed] [Google Scholar]
  • 21.Crémieux PY, Ledoux S, Clerici C, Cremieux F, Buessing M. The impact of bariatric surgery on comorbidities and medication use among obese patients. Obes Surg. 2010;20(7):861-870. doi: 10.1007/s11695-010-0163-6 [DOI] [PubMed] [Google Scholar]
  • 22.Finkelstein EA, Allaire BT, Burgess SM, Hale BC. Financial implications of coverage for laparoscopic adjustable gastric banding. Surg Obes Relat Dis. 2011;7(3):295-303. doi: 10.1016/j.soard.2010.10.011 [DOI] [PubMed] [Google Scholar]
  • 23.Gesquiere I, Aron-Wisnewsky J, Foulon V, et al. Medication cost is significantly reduced after Roux-en-Y gastric bypass in obese patients. Obes Surg. 2014;24(11):1896-1903. doi: 10.1007/s11695-014-1325-8 [DOI] [PubMed] [Google Scholar]
  • 24.Hodo DM, Waller JL, Martindale RG, Fick DM. Medication use after bariatric surgery in a managed care cohort. Surg Obes Relat Dis. 2008;4(5):601-607. doi: 10.1016/j.soard.2008.01.008 [DOI] [PubMed] [Google Scholar]
  • 25.Karim MA, Clifton E, Ahmed J, Mackay GW, Ali A. Economic evaluation of bariatric surgery to combat morbid obesity: a study from the West of Scotland. Asian J Endosc Surg. 2013;6(3):197-202. doi: 10.1111/ases.12042 [DOI] [PubMed] [Google Scholar]
  • 26.Morgan DJ, Ho KM, Armstrong J, Litton E. Long-term clinical outcomes and health care utilization after bariatric surgery: a population-based study. Ann Surg. 2015;262(1):86-92. doi: 10.1097/SLA.0000000000000972 [DOI] [PubMed] [Google Scholar]
  • 27.Mullen DM, Marr TJ. Longitudinal cost experience for gastric bypass patients. Surg Obes Relat Dis. 2010;6(3):243-248. doi: 10.1016/j.soard.2010.01.002 [DOI] [PubMed] [Google Scholar]
  • 28.Sussenbach SP, Padoin AV, Silva EN, et al. Economic benefits of bariatric surgery. Obes Surg. 2012;22(2):266-270. doi: 10.1007/s11695-011-0558-z [DOI] [PubMed] [Google Scholar]
  • 29.Christou NV, Sampalis JS, Liberman M, et al. Surgery decreases long-term mortality, morbidity, and health care use in morbidly obese patients. Ann Surg. 2004;240(3):416-423. doi: 10.1097/01.sla.0000137343.63376.19 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Sampalis JS, Liberman M, Auger S, Christou NV. The impact of weight reduction surgery on health-care costs in morbidly obese patients. Obes Surg. 2004;14(7):939-947. doi: 10.1381/0960892041719662 [DOI] [PubMed] [Google Scholar]
  • 31.Finkelstein EA, Allaire BT, DiBonaventura MD, Burgess SM. Direct and indirect costs and potential cost savings of laparoscopic adjustable gastric banding among obese patients with diabetes. J Occup Environ Med. 2011;53(9):1025-1029. doi: 10.1097/JOM.0b013e318229aae4 [DOI] [PubMed] [Google Scholar]
  • 32.Klein S, Ghosh A, Cremieux PY, Eapen S, McGavock TJ. Economic impact of the clinical benefits of bariatric surgery in diabetes patients with BMI ≥35 kg/m2. Obesity (Silver Spring). 2011;19(3):581-587. doi: 10.1038/oby.2010.199 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Makary MA, Clark JM, Shore AD, et al. Medication utilization and annual health care costs in patients with type 2 diabetes mellitus before and after bariatric surgery. Arch Surg. 2010;145(8):726-731. doi: 10.1001/archsurg.2010.150 [DOI] [PubMed] [Google Scholar]
  • 34.Statistics Canada Population estimates, quarterly. Table 17-10-0009-01. Accessed June 17, 2020. https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1710000901
  • 35.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Prev Med. 2007;45(4):247-251. doi: 10.1016/j.ypmed.2007.08.012 [DOI] [PubMed] [Google Scholar]
  • 36.Saleh F, Doumouras AG, Gmora S, Anvari M, Hong D. Outcomes the Ontario Bariatric Network: a cohort study. CMAJ Open. 2016;4(3):E383-E389. doi: 10.9778/cmajo.20150112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Honarmand K, Chetty K, Vanniyasingam T, Anvari M, Chetty VT. Type 2 diabetes remission rates 1-year post-Roux-en-Y gastric bypass and validation of the DiaRem score: the Ontario Bariatric Network experience. Clin Obes. 2017;7(3):176-182. doi: 10.1111/cob.12189 [DOI] [PubMed] [Google Scholar]
  • 38.Tarride JE, Breau R, Sharma AM, et al. The effect of bariatric surgery on mobility, health-related quality of life, healthcare resource utilization, and employment status. Obes Surg. 2017;27(2):349-356. doi: 10.1007/s11695-016-2298-6 [DOI] [PubMed] [Google Scholar]
  • 39.Doumouras AG, Lee Y, Babe G, et al. The hidden cost of an extensive preoperative work-up: predictors of attrition after referral for bariatric surgery in a universal healthcare system. Surg Endosc. 2020;34(2):988-995. doi: 10.1007/s00464-019-06894-9 [DOI] [PubMed] [Google Scholar]
  • 40.Hunt KD, Doumouras AG, Lee Y, Gmora S, Anvari M, Hong D. The effect of surrogate procedure volume on bariatric surgery outcomes: do common laparoscopic general surgery procedures matter? Surg Endosc. 2020;34(3):1278-1284. doi: 10.1007/s00464-019-06897-6 [DOI] [PubMed] [Google Scholar]
  • 41.Lovrics O, Doumouras AG, Gmora S, Anvari M, Hong D. Metabolic outcomes after bariatric surgery for Indigenous patients in Ontario. Surg Obes Relat Dis. 2019;15(8):1340-1347. doi: 10.1016/j.soard.2019.05.035 [DOI] [PubMed] [Google Scholar]
  • 42.Tu K, Mitiku TF, Ivers NM, et al. Evaluation of Electronic Medical Record Administrative Data Linked Database (EMRALD). Am J Manag Care. 2014;20(1):e15-e21. [PubMed] [Google Scholar]
  • 43.Orr SK, Tu K, Carsley S, et al. Using EMRALD to assess baseline body mass index among children living within and outside communities participating in the Ontario, Canada Healthy Kids Community Challenge. PLoS One. 2019;14(4):e0213443. doi: 10.1371/journal.pone.0213443 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Wodchis WPBK, Nikitovic M, McKillop I Guidelines on person-level costing using administrative databases in Ontario. Working paper series. Health System Performance Research Network; 2013. Accessed June 17, 2020. https://tspace.library.utoronto.ca/bitstream/1807/87373/1/Wodchis%20et%20al_2013_Guidelines%20on%20Person-Level%20Costing.pdf
  • 45.Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424. doi: 10.1080/00273171.2011.568786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Stürmer T, Wyss R, Glynn RJ, Brookhart MA. Propensity scores for confounder adjustment when assessing the effects of medical interventions using nonexperimental study designs. J Intern Med. 2014;275(6):570-580. doi: 10.1111/joim.12197 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Lewis KH, Zhang F, Arterburn DE, Ross-Degnan D, Gillman MW, Wharam JF. Comparing medical costs and use after laparoscopic adjustable gastric banding and Roux-en-Y gastric bypass. JAMA Surg. 2015;150(8):787-794. doi: 10.1001/jamasurg.2015.1081 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Spaniolas K, Goldberg I, Yang J, et al. Hospital utilization 4 years after bariatric surgery: sleeve gastrectomy versus Roux-en-Y gastric bypass. Surg Obes Relat Dis. 2019;15(9):1465-1472. doi: 10.1016/j.soard.2019.06.023 [DOI] [PubMed] [Google Scholar]
  • 49.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28(25):3083-3107. doi: 10.1002/sim.3697 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Colla CH, Wennberg DE, Meara E, et al. Spending differences associated with the Medicare Physician Group Practice Demonstration. JAMA. 2012;308(10):1015-1023. doi: 10.1001/2012.jama.10812 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Dimick JB, Ryan AM. Methods for evaluating changes in health care policy: the difference-in-differences approach. JAMA. 2014;312(22):2401-2402. doi: 10.1001/jama.2014.16153 [DOI] [PubMed] [Google Scholar]
  • 52.Joo H, Dunet DO, Fang J, Wang G. Cost of informal caregiving associated with stroke among the elderly in the United States. Neurology. 2014;83(20):1831-1837. doi: 10.1212/WNL.0000000000000986 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Joynt KE, Chan DC, Zheng J, Orav EJ, Jha AK. The impact of Massachusetts health care reform on access, quality, and costs of care for the already-insured. Health Serv Res. 2015;50(2):599-613. doi: 10.1111/1475-6773.12228 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Rosenthal MB, Alidina S, Friedberg MW, et al. A difference-in-difference analysis of changes in quality, utilization and cost following the Colorado multi-payer patient-centered medical home pilot. J Gen Intern Med. 2016;31(3):289-296. doi: 10.1007/s11606-015-3521-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Wing C, Simon K, Bello-Gomez RA. Designing difference in difference studies: best practices for public health policy research. Annu Rev Public Health. 2018;39:453-469. doi: 10.1146/annurev-publhealth-040617-013507 [DOI] [PubMed] [Google Scholar]
  • 56.Efron BTR. An Introduction to the Bootstrap. Chapman & Hall; 1993. doi: 10.1007/978-1-4899-4541-9 [DOI] [Google Scholar]
  • 57.Gregori D, Petrinco M, Bo S, Desideri A, Merletti F, Pagano E. Regression models for analyzing costs and their determinants in health care: an introductory review. Int J Qual Health Care. 2011;23(3):331-341. doi: 10.1093/intqhc/mzr010 [DOI] [PubMed] [Google Scholar]
  • 58.Finkelstein EA, Allaire BT, Globe D, Dixon JB. The business case for bariatric surgery revisited: a non-randomized case-control study. PLoS One. 2013;8(9):e75498. doi: 10.1371/journal.pone.0075498 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Maciejewski ML, Livingston EH, Smith VA, Kahwati LC, Henderson WG, Arterburn DE. Health expenditures among high-risk patients after gastric bypass and matched controls. Arch Surg. 2012;147(7):633-640. doi: 10.1001/archsurg.2012.818 [DOI] [PubMed] [Google Scholar]
  • 60.Bruschi Kelles SM, Machado CJ, Barreto SM. Before-and-after study: does bariatric surgery reduce healthcare utilization and related costs among operated patients? Int J Technol Assess Health Care. 2015;31(6):407-413. doi: 10.1017/S0266462315000653 [DOI] [PubMed] [Google Scholar]
  • 61.Rios-Diaz AJ, Metcalfe D, Devin CL, Berger A, Palazzo F. Six-month readmissions after bariatric surgery: results of a nationwide analysis. Surgery. 2019;166(5):926-933. doi: 10.1016/j.surg.2019.06.003 [DOI] [PubMed] [Google Scholar]
  • 62.Morton JM, Ponce J, Malangone-Monaco E, Nguyen N. Association of bariatric surgery and national medication use. J Am Coll Surg. 2019;228(2):171-179. doi: 10.1016/j.jamcollsurg.2018.10.021 [DOI] [PubMed] [Google Scholar]
  • 63.Hachem A, Brennan L. Quality of life outcomes of bariatric surgery: a systematic review. Obes Surg. 2016;26(2):395-409. doi: 10.1007/s11695-015-1940-z [DOI] [PubMed] [Google Scholar]
  • 64.Larsen AT, Højgaard B, Ibsen R, Kjellberg J. The socio-economic impact of bariatric surgery. Obes Surg. 2018;28(2):338-348. doi: 10.1007/s11695-017-2834-z [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement.

eTable 1. Administrative Data Sources

eFigure 1. Study Flow Diagram

eTable 2. Comparisons of Unmatched and Matched RYGB Cohorts

eTable 3. Comparisons of Unmatched and Matched Controls

eTable 4. Total Healthcare Expenditures and Individual Cost Components for Matched RYGB Patients, Per Year of Analysis

eTable 5. Total Healthcare Expenditures and Individual Cost Components for Matched Controls, Per Year of Analysis

eTable 6. Difference-in-Differences (DID) in Healthcare Expenditures and Resource Utilization for Patients With Complete Data at Month 60

eFigure 2. Costs Before and After Surgery for Unmatched and Matched RYGB Patients

Click here for additional data file. (334.6KB, pdf)

Articles from JAMA Surgery are provided here courtesy of American Medical Association

RESOURCES