Predictors of Postoperative Aftercare Attrition among Gastric Bypass Patients

Abstract

Background: Poor adherence to post-bariatric surgery aftercare continues to challenge surgical practices. The objective of this study was to identify factors that predict poor aftercare attendance among patients who underwent Roux-en-Y Gastric Bypass (RYGB) surgery.

Method: A retrospective medical chart review of patients who underwent RYGB from 2002 to 2011 was conducted. Patients with four visits or more in the first 2 years (>50%) were categorized as “acceptable follow-up” and with ≤50% as “poor follow-up.” Demographics, presurgical body mass index (BMI), and comorbidities were compared using multivariate analysis.

Results: Out of 2,658 patients, 1,092 (41.1%) had acceptable follow-up. Preoperative factors that predicted acceptable follow-up included female gender (odds ratio [OR] 1.41 [95% confidence interval (CI) 1.15–1.72]), older age (OR 1.03 [95% CI 1.03–1.04]), higher BMI at surgery (OR 1.02 [95% CI 1.01–1.03]), and Hispanic ethnicity (OR 1.40 [95% CI 1.15–1.72]). Conversely, presence of diabetes mellitus (OR 0.58 [95% CI 0.39–0.88]), hypertension (OR 0.53 [95% CI 0.39–0.72]), and obstructive sleep apnea (OR 0.39 [95% CI 0.26–0.57]) predicted less adherence to RYGB aftercare.

Conclusion: These findings suggest RYGB patients' age, gender, ethnicity, preoperative BMI, and certain comorbidities should be considered to maximize postoperative aftercare attendance.

Introduction

Bariatric surgery is a well-established treatment option for severe obesity and its related comorbidities.^1–3 However, bariatric surgery must be considered as a “tool” for permanent weight loss and not the end-all solution. Specifically, while the surgical procedure is an important component of a comprehensive weight loss program, it is one component among many and is couched in a multistep, multidisciplinary, comprehensive care model. Ideally, this model begins with orientation and education of patients before surgery and continues with lifelong postoperative aftercare to realize maximum weight loss and health benefits.

Postoperative aftercare or follow-up visits are important after both restrictive and malabsorptive procedures. Percentage of excess weight loss after adjustable gastric band surgery has been shown to be significantly related to the adherence to postoperative aftercare.^4–7 Patients who undergo procedures with a malabsorptive component (including Roux-en-Y Gastric Bypass [RYGB] and biliopancreatic diversion) are at higher risk of nutritional deficiencies including deficiencies of protein, vitamins, and micronutrients.^8–10 As such, consistent postoperative follow-up in this group of patients is necessary to prevent nutritional deficiencies and assure appropriate weight loss.^11,12 Moreover, poor retention in postoperative bariatric aftercare has detrimental effects on patient safety due to delayed diagnosis of complications and also on unbiased and internal and external research study results.¹³

Most bariatric surgery centers have standardized patient postoperative aftercare protocols, but regardless, patient attrition is a consistent problem in the field, particularly beyond 1 year after surgery.¹⁴ The literature reports that follow-up rates can be as low as 50% at 1 year,^12,15 30% at 2 years, and <10% at 10 years after surgery.¹⁶ Even national databases such as the Bariatric Outcomes Longitudinal Database that includes hundreds of centers of surgical excellence from all states in the nation except Vermont and New Mexico suffer from long-term poor follow-up rates.^17,18

Improving postoperative follow-up attrition rates, and potentially the long-term health outcomes in bariatric surgery patients, requires the identification of patient characteristics that predict both satisfactory and poor follow-up. By identifying specific patient characteristics, surgery programs could incorporate specific strategies to improve attendance at surgical aftercare programs. Therefore, this study aimed to identify preoperative patient characteristics that predict postoperative adequate and poor post-RYGB aftercare follow-up.

Patients and Methods

A retrospective database review of 2,658 adult morbidly obese patients who underwent RYGB by a single surgeon from 2002 to 2011 was analyzed. This study was approved by the University of Miami Institutional Review Board. Data were collected on patient demographics (age, gender, and ethnicity), presurgical body mass index (BMI), existing comorbidities, and follow-up visits at each scheduled time point.

Surgical eligibility and procedure

All selected patients met the National Institutes of Health criteria for bariatric surgery¹⁹: either a BMI >40 kg/m² or a BMI >35 kg/m² with at least one comorbidity. All patients underwent group and individual education with the surgeon and an interdisciplinary group of paraprofessionals (e.g., dietician, registered nurse). All surgical and nonsurgical options, outcomes, complications, and necessary lifestyle changes were discussed. All patients underwent psychological and nutritional evaluations before surgery. The laparoscopic approach was used in all but 14 patients. Patients underwent RYGB in an antecolic, antegastric manner.²⁰

Follow-up

Patients were visited by the registered nurse at 1 week. Planned follow-up visits with the surgeon were scheduled at 1, 3, 6, 12, 18, and 24 months post-surgery (six visits in total) and thereafter annually. Patients attending four or more visits with the surgeon in the first 2 years (>50%) were categorized as “acceptable follow-up” and those attending ≤50% of visits with the surgeon were classed as “poor follow-up.” Age, gender, ethnicity, presurgical BMI, social history, and comorbidities were compared between two groups.

Statistical analysis

Baseline characteristics were compared using chi-square for categorical variables, and analysis of variance or Mann–Whitney U-tests for continuous variables, as appropriate, using SPSS v18.0.0 (SPSS, Inc., Chicago, IL). A logistic regression model was fit to examine the contribution of each covariate to acceptable (or nonacceptable) follow-up. Specifically, the model was fit as a binary outcome (Y=acceptable follow-up; N=poor follow-up), and included patient characteristics as covariates with p<0.1 found in univariate analysis. Adjustments were made in a stepwise backward elimination procedure, and odds ratios (OR) were reported with corresponding 95% confidence intervals (CI). Statistical tests resulting in a probability of ≤0.05 were considered statistically significant.

Results

Our sample consisted of 2,658 patients (77% female) with a mean age of 41.2±12.5 years (range 18–78 years). Follow-up rates at 1, 3, 6, 12, 18, and 24 months after surgery were 84.5%, 60%, 54.6%, 53.9%, 29%, and 28.6%, respectively. A total of 1,566 (58.9%) patients were classified as having poor follow-up (≤50% planned visits) and 1,092 (41.1%) had acceptable follow-up. Patients with poor follow-up were younger (39.9±11.9 years vs. 43±12.5 years) and more likely to be male (25.5% vs. 20.1%) than patients with acceptable follow-up visits. There were more smokers, more non-Hispanic blacks, and fewer Hispanics in the poor follow-up group. The poor follow-up group had a higher frequency of comorbidities, especially diabetes mellitus, hypertension, congestive heart failure, hyperlipidemia, diagnosed fatty liver, asthma, obstructive sleep apnea, depression, gastroesophageal reflux disease (requiring medical therapy), and stress urinary incontinence, compared with the acceptable follow-up group (Table 1).

Table 1.

Comparison of Factors Between Roux-en-Y Gastric Bypass Patients with Acceptable Aftercare Follow-Up and Poor Follow-Up

	Acceptable follow-up (n=1,092)	Poor follow-up (n=1,566)	p-Value
Age, mean±SD, years	43.3±12.5	39.9±11.9	<0.001
Female, n (%)	873 (79.9%)	1166 (74.5%)	0.001
Ethnicity, n (%)
Non-Hispanic black	176 (16.1%)	339 (21.6%)	<0.001
Non-Hispanic white	350 (32.1%)	519 (33.1%)
Hispanic	566 (51.8%)	708 (45.2%)
Initial BMI, mean±SD, kg/m2	47.3±8.0	46.9±8.1	0.062
Social history
Substance abuse	3 (0.3%)	5 (0.3%)	0.873
Smoking	20 (1.8%)	55 (3.5%)	0.010
Alcohol	33 (3.0%)	68 (4.3%)	0.080
Comorbid conditions
Ischemic heart disease	6 (0.5%)	15 (1.0%)	0.242
Congestive heart failure	4 (0.4%)	19 (1.2%)	0.020
History of stroke/TIA	2 (0.2%)	6 (0.4%)	0.483
Diabetes mellitus	44 (4.0%)	149 (9.5%)	<0.001
Hypertension	92 (8.4%)	271 (17.3)	<0.001
Hyperlipidemia	59 (5.4%)	154 (9.8%)	<0.001
Gout/hyperuricemia	10 (0.9%)	23 (1.5%)	0.205
Fatty liver (diagnosed)	28 (2.6%)	74 (4.7%)	0.004
Asthma	24 (2.2%)	74 (4.7%)	0.001
Obstructive sleep apnea	39 (3.6%)	194 (12.4%)	<0.001
Depression	63 (5.8%)	163 (10.4%)	<0.001
GERD on medication	33 (3.0%)	97 (6.2%)	<0.001
Stress urinary incontinence	37 (3.4%)	114 (7.3%)	<0.001

SD, standard deviation; BMI, body mass index; TIA, transient ischemic attack; GERD, gastroesophageal reflux disease.

Multivariate analysis found the preoperative factors that independently predicted acceptable follow-up, which included female gender (odds ratio [OR] 1.41 [95% confidence interval (CI) 1.15–1.72]), older age (OR 1.03 [95% CI 1.03–1.04]), higher BMI at surgery (OR 1.02 [95% CI 1.01–1.03]), and Hispanic ethnicity (OR 1.40 [95% CI 1.15–1.72]). Conversely, presence of diabetes mellitus (OR 0.58 [95% CI 0.39–0.88]), hypertension (OR 0.53 [95% CI 0.39–0.72]), and obstructive sleep apnea (OR 0.39 [95% CI 0.26–0.57]) predicted less adherence to RYGB aftercare (Table 2).

Table 2.

Multivariate Logistic Regression for Predictors of Acceptable Follow-Up After Roux-en-Y Gastric Bypass

Patient characteristics	B	SE	Wald	p-Value	OR	CI 95%
Age	0.032	0.004	81.454	<0.001	1.033	1.025–1.040
Female gender	0.342	0.103	11.071	0.001	1.408	1.151–1.722
Ethnicity*
Non-Hispanic black	−0.138	0.123	1.268	0.260	0.871	0.685–1.108
Hispanic	0.337	0.094	12.886	<0.001	1.400	1.165–1.683
BMI at surgery	0.017	0.005	9.828	0.002	1.017	1.006–1.027
Diabetes mellitus	−0.539	0.209	6.667	0.010	0.584	0.388–0.878
Hypertension	−0.631	0.155	16.503	<0.001	0.532	0.393–0.722
Obstructive sleep apnea	−0.954	0.199	22.649	<0.001	0.389	0.263–0.574
Stress urinary incontinence	−0.413	0.215	3.684	0.055	0.662	0.434–1.009

^*Non-Hispanic white.

SE, standard error; OR, odds ratio; CI, confidence interval.

Discussion

Our analysis suggests that RYGB patients' age, gender, ethnicity, preoperative BMI, and certain comorbidities should be considered to maximize postoperative aftercare attendance. Younger men with a lower BMI and more comorbidities may require more intensive follow-up efforts, specifically.

Our analysis showed that the follow-up rate at 1 and 2 years after RYGB was 53.9% and 28.6%, respectively. Well in excess of half of patients (58.9%) missed at least three out of six planned visits in the first 2 years after surgery, which is similar to other reports. Inabnet et al.¹⁷ published an analysis of 186,576 research-consented patients in the Bariatric Outcomes Longitudinal Database. In their analysis, data were available only for 50% of patients at 1 year after surgery. Higa et al.¹⁶reported a 33% office follow-up rate at 2 years and 7% at 10 years after RYGB. Lennerz et al.²¹ also reported a follow-up rate of <30% at 2 years in 345 adolescents and young adults (up to 21 years of age) who underwent bariatric surgery in Germany. All of these studies document the consistent challenges in retention of bariatric patients in postoperative follow-ups, even as bariatric surgery becomes more mainstream and acceptable.

Certainly, poor attendance at follow-up appointments may lead to less compliance with postoperative instructions and inadequate weight loss after RYGB.^11,12,22 Longitudinal follow-up may help with weight loss and its maintenance by reinforcing dietary instruction, educating and counseling patients with inappropriate food choices or eating behavior, individualizing realistic weight loss goals, encouraging patients to pursue these goals, and holding them responsible for maintaining the healthy lifestyle changes.¹¹ Attending support group meetings has also been shown to be associated with greater weight loss after RYGB.²³ Thus, long-term multidisciplinary care seems to be an important component in RYGB aftercare; and appropriate follow-up retention strategies should be utilized to encourage patients to continue to attend their bariatric appointments. Gourash et al.¹³ presented the result of extensive retention strategies for a subgroup of patients in the Longitudinal Assessment of Bariatric Surgery study to maximize data completeness. They utilized flexible scheduling, a call protocol, reminder letters, abbreviated visit options, honoraria, travel reimbursement, newsletters, study Web site, retention surveys, frequent updating of contact information, sending registered letters, and searching medical and public records. They obtained weight information for 95.2% and 92.2% patients at 1 and 2 years after surgery, respectively. However, the majority of surgery clinics cannot afford or sustain these required resources (budget and manpower) to implement all of these strategies for all patients. Other strategies that are not as resource intensive such as telephone follow-up have been used by some centers and may be useful to gather weight information, but they are not as effective or accurate as face-to-face encounter for ensuring adherence of patients to required diet and lifestyle changes.²³ Consequently, a feasible way to address this problem could be a selective approach for a subgroup of patients who are at higher risk for attrition. This group of patients should be recognized using predictors of attrition and followed vigorously using several strategies.

In the present study, older age predicted better follow-up attendance. Wheeler et al.²⁴ also found that older bariatric patients were more likely to adhere to postoperative follow-up. This may be due to the higher importance older people place on their personal health. Older patients may also have more stable insurance coverage that facilitates their more consistent follow-up attendance. Conversely, younger patients may be busier in the household with childcare, job demands, and/or social activities. Some studies failed to show the association between age and attendance to postoperative follow-up.²⁵ This may be explained by low volume of patients in some studies, different definitions of nonattendance, and different age ranges that may not include both extremes of young and old ages.

In the present analysis, female gender predicted better attendance at follow-ups, which has not been consistently shown in other published studies. Toussi et al.²⁶ similarly showed women had more attendance at 2 years after RYGB. The other factor that significantly predicted adherence to post-bariatric surgery aftercare in the current sample was a higher BMI at surgery. Patients with a higher BMI may feel the need to be more closely monitored and may require the extra social support provided by attending postoperative visits for support.

Hispanic ethnicity was also significantly associated with better follow-up rates in the present analysis. However, this finding may be a local phenomenon, as in the region of this study (South Florida), Hispanics are the majority of the population. All staff, including the surgeon in the outpatient department and the majority of hospital staff, are Hispanic. Thus, local medical staff who are involved in patients' aftercare can effectively communicate with Hispanic patients, encourage them to consistently follow postoperative instructions, and discuss their concerns in detail, all in Spanish if needed. This finding implies the possible risk of attrition in minority groups where there is not ethnic concordance between healthcare staff and patients, particularly if there is a language barrier. Minority groups may suffer from poor interaction and inefficient communication with healthcare staff.²⁷ This may affect efficacy of preoperative and postoperative educations and success of motivating patients during aftercare. Conversely, being of the same ethnicity can enhance communication and interpersonal interactions by common cultural backgrounds and reduction of language barriers.

One of the unique findings of the current study showed that diabetes mellitus, hypertension, and obstructive sleep apnea all independently predict attrition. Since comorbid conditions usually require their own follow-ups, these patients usually visit their primary physician or other healthcare providers. They may undergo serial medical visits including lab tests and nutrition counseling by family physicians, internists, endocrinologists, or other specialists. Thus, these patients may be less motivated to attend bariatric surgery clinic, especially if they do not experience any complication related to their bariatric surgery. However, an exact explanation of this finding needs further examination by extracting information from patients' appointments with their other physician specialists.

Several other factors have been shown to be associated with aftercare attrition or nonadherence in other studies. These factors include greater travel distance to the follow-up clinic,^7,25,28 avoidant relationship style,²⁸ higher scores in Beck Depression Inventory, having a diagnosis of depression or any psychological disorder,²⁶ binge eating, personality disorders, environmental stressors,²² not having insurance coverage, being unemployed, and being married.²⁴ These findings, along with the aforementioned research, show that aftercare attrition is multifactorial and is rooted in a variety of complex somatic, psychological, and social issues. Moving from the geographic area, switching to other bariatric centers after facing a complication, being managed by other healthcare teams other than the patient's bariatric team, or self-management of supplements without any medical follow-up can all contribute to poor follow-up in aftercare.

Study limitations

Although this analysis is distinctive due to the large sample size and inclusion of pre-existing patient comorbidities, there are study limitations that should be mentioned. Mainly, due to this being a retrospective database review, not all potential predictors of postoperative follow-up that a prospective study would allow could be extracted, such as psychosocial factors, travel distance to the outpatient clinic, and other appointments with other healthcare providers. Also, the results cannot be generalized to other populations, since the study has carried out in south Florida with specific ethnic subgroups. Other studies that investigated factors contributing to postoperative attrition have been similarly limited because of small number of studies, varieties in types of bariatric surgery, different definitions of attrition and methodologies, and different type of variables considered.¹⁵

Conclusion

Analysis here showed that patients' ethnicity, gender, age, BMI, and preoperative comorbidities should be considered to maximize postoperative aftercare attendance. In particular, younger men with lower BMI, those with more comorbidities, and minority ethnic groups may require more intensive follow-up efforts.

Author Disclosure Statement

No competing financial interests exist.