Abstract
Abstract
Background:
Hospital readmissions following bariatric surgery are high and it is necessary to identify modifiable risk factors to minimize this postoperative cost. We hypothesize that lower levels of education and health literacy are associated with increased risks of nonadherence, thus leading to increased emergency department (ED) visits and preventable readmissions postoperatively.
Methods:
Bariatric surgery patients presenting between October 2015 and December 2016 were administered a preoperative questionnaire that measured education level and the Rapid Estimate of Adult Literacy in Medicine—Short Form (REALM-SF) health literacy test. The rates of postoperative ED visits and readmissions were across education levels (≤12th grade versus >12th grade) and health literacy scores (≤8th grade versus high school level). A composite “hospital visit” outcome was also assessed.
Results:
Ninety-five patients were enrolled; 23 had ≤12th grade level education and 7 scored ≤8th grade on the REALM-SF. Patients with ≤12th grade education were significantly more likely to have a hospital visit after surgery, compared with patients with >12th grade education (incidence rate ratio [IRR] 3.06, P = .008). No significant difference in ED visits, readmission, or hospital visits was seen when stratified by REALM-SF health literacy score.
Conclusions:
Lower level of education was associated with more than threefold increased risk of postoperative ED visits and readmission in our center's bariatric surgery patients. A patient's education level is a low-cost means to identify patients who are at risk for postoperative hospital visits, and who may benefit from enhanced educational efforts or more intensive postoperative follow-up.
Introduction
O
In bariatric surgery, postoperative emergency department (ED) visits and readmission are common and may be a significant contributor to cost. Encinosa et al. found an 18% rate of readmission or ED visit within the first 180 days after bariatric surgery and that a readmission increased the 180-day cost of bariatric surgery from ∼$27,000 to $65,000. 4 Prior analysis of the etiology of postoperative ED visits in bariatric surgery has found that many of these ED visits may be avoidable. Kellogg et al. found that of 173 postbariatric surgery ED visits, 54% were attributed to nausea/vomiting/dehydration, benign abdominal pain, or wound issues (26%, 20%, and 8%, respectively). 5 A similar study of 42 readmissions following laparoscopic Roux-en-Y gastric bypass found 57% were due to these same complaints. 6 Some socioeconomic factors have been found to be correlated with a higher rate of ED visits and readmission. Several studies have shown statistically significant higher rates of ED visits and readmission in Medicare and Medicaid patients after bariatric surgery.6–9 Patients who are unemployed, disabled, or retired have also been found to have higher rates of postoperative ED visit or readmission. 5
The importance of the negative impact of postoperative readmission on cost and patient satisfaction is underscored by the inclusion of readmission recently as a hospital quality measure. The Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQUIP) requires tracking of readmissions by its accredited inpatient centers. In 2015, the MBSAQUIP launched a program with a goal of decreasing bariatric readmissions through a bundled process. 10 Financial incentives have also been implemented as a means to encourage hospitals to decrease rate of readmission. In 2012, the Centers for Medicare and Medicaid Services (CMS) proposed a plan for negative payment adjustments on hospitals with excessive readmissions. 11
Further understanding of risk factors for ED visits and readmissions after bariatric surgery is needed. While several socioeconomic factors, including Medicaid insurance status, have been correlated, the effects of health literacy and education level on ED visits and readmission have not been well studied in the literature. We hypothesize that patients with less formal education and lower healthcare literacy are more likely to have an incomplete understanding of postoperative instructions and appropriate milestones, and would, therefore, be at higher risk of postoperative ED visits and readmission. Moreover, if health literacy and education are relatively easy to measure and can provide an opportunity to identify high-risk patients, this would allow clinicians to individualize patient education and follow-up to optimize outcomes.
Materials and Methods
Patient recruitment
All patients undergoing bariatric surgery at our center from October 1, 2015, to December 27, 2016, were offered voluntary participation in this study at their final preoperative clinic visit, and followed prospectively. Five patients who were initially enrolled did not go on to have bariatric surgery and were ineligible for the study. Patient follow-up was administratively censored on December 27, 2016, and patients with <30 days follow-up were excluded from analysis. Institutional Review Board (IRB) approval was obtained from the University of North Carolina.
Education and health literacy testing
During the final preoperative clinic visit, enrolled patients completed a questionnaire. Patient education level was interpreted and dichotomized as ≤12 years of education (those without a high school degree or those with no post-high school education) and >12 years of education (those with post-high school education). The general equivalency diploma was considered the equivalent of a high school degree.
Enrolled patients were also administered the Rapid Estimate of Adult Literacy in Medicine—Short Form (REALM-SF) assessment during the same visit to measure health literacy. Briefly, the REALM-SF is a 3–4-minute proctor-administered assessment. The patient is asked to pronounce seven health-related words, and the number appropriately pronounced is scored. The maximum score of 7 suggests high school level health literacy, and that the patient would be able to read and understand most instruction materials. A score <7 implies less than high school level of health literacy and suggests the patient would benefit from low-literacy materials. This test has been validated with three other standardized reading tests, including the Peabody Individual Achievement Test—Revised (PIAT-R), the Wide Range Achievement Test—Revised (WRAT-R), and the Slosson Oral Reading Test—Revised (SORT-R). 12 For this study, the REALM-SF was chosen for its ease of use, free access, and validation compared with more time-consuming literacy tests. Patient literacy scores were also dichotomized into high school education (a score of 7) and less than a high school education (a score ≤6).
Data collection
The primary outcomes of interest included rate of postoperative ED visits (i.e., a discharge from the ED without admission) and hospital readmissions. Readmission includes observation and inpatient readmissions. A composite hospital visit variable, which included the sum of ED visits and readmissions, was also created. Only a patient's first ED visit and readmission, respectively, were included in the analysis. Counseling from a bariatric nurse navigator or an unplanned postoperative clinic visit was the existing alternative to ED visit and not included in analysis as these were felt to be lower cost alternatives to ED visit and readmission.
Data were collected from review of the electronic medical record (EMR). ED visits and readmissions occurring at our center, as well as those occurring at other centers and documented in our center's EMR, were included. Diagnoses and therapeutic interventions were recorded for each ED visit and readmission. ED visits and readmissions that were seemingly unrelated to bariatric surgery (e.g., a fall resulting in a knee injury several months after surgery) were excluded from the analysis. In addition, patient age, sex, body mass index, excess body weight, procedure performed (laparoscopic sleeve gastrectomy or laparoscopic Roux-en-Y gastric bypass), surgical complications, and preoperative use of antihypertensives, diabetes medications, or insulin were abstracted.
Statistical analyses
Patient demographics and preoperative medication usage were described using univariate analyses. Potential differences in patient demographics and medications were compared across health literacy score and education level using Fisher's exact and Student's t-tests, where appropriate. A P value <.05 was considered significant.
The rate of ED visits, hospital readmissions, and hospital visits per 100 days was compared across health literacy and education levels using Poisson regression. Patients with high education and health literacy were used as reference. The incidence of surgical complications across health literacy and education level was compared using Fisher's exact tests. All analyses were performed using SAS 9.4 (SAS, Inc., Cary, NC).
Results
Overall, 100 of 106 eligible patients (94.3%) screened agreed to participate in the study. Ninety-five patients had at least 30 days of follow-up and were included in the analyses. Average follow-up was 252 days (standard deviation 98.4). Only 7 patients (7.5%) had a health literacy score ≤6, and 23 (24.5%) had a high school education or less. See Table 1 for patient demographics.
≤12 years includes “some high school,” “GED,” and “12th grade”; >12 years includes “some college,” “Bachelor's degree,” and “graduate degree.”
One laparoscopic sleeve was converted to open surgery.
SD, standard deviation; GED, general equivalency diploma; RYGB, Roux-en-Y gastric bypass.
No significant differences in gender (P = .08), age (P = .82), excess weight (P = .29), hypertension medication (P = .70), oral diabetes medication (P = .99), insulin (P = .99), or procedure type (P = .99) were seen across health literacy score. Similarly, when compared across education level, no significant differences across gender (P = .51), age (P = .73), excess weight (P = .11), hypertension medication (P = .15), oral diabetes medication (P = .43), insulin (P = .73), or procedure type (P = .99) were seen.
There were 10 complications of surgery, including surgical-site infection, deep vein thrombosis, superior mesenteric vein thrombosis, anastomotic stricture, marginal ulcer, and gastric torsion. Complication rate was higher in patients with lower education (17.4% versus 8.5%, P = .25) and lower level of health literacy (28.6% versus 9.1%, P = .16), but no statistically significant differences were seen.
A total of 26.3% (25/95) of patients had a postoperative hospital visit. Patients with a high school education or less were significantly more likely to have a hospital visit after surgery (incidence rate ratio [IRR] 3.06, 95% CI 1.39–6.73, P = .008) (Table 2). While increased rates of ED visits (0.11 per 100 days versus 0.04 per 100 days) and readmissions (0.13 per 100 days and 0.04 per 100 days) were also seen, these differences were not statistically significant, P = .15 and P = .07. Moreover, as patient education level increased, the number of patients with a hospital visit decreased (Fig. 1A, B).
Bold value is statistically significant.
P-values were calculated using Poisson regression; a P-value <.05 was considered significant.
Includes ED visits and readmissions.
CI, confidence interval; ED, emergency department; IRR, incidence rate ratio.
When analyzed by REALM-SF score, patients with low health literacy did not have significant differences in ED visits (IRR 0.87, P = .90), readmissions (IRR 2.18, P = .35), or overall hospital visits (IRR 1.56, P = .49) when compared with patients with high health literacy (Table 2).
Among all ED visits, 16/19 (84%) carried diagnoses of abdominal pain, nausea, dehydration, constipation, or gastric reflux with no significant anatomical or laboratory abnormalities identified on work-up, and resulted in ED discharge with supportive care. The remainder of ED visits included a surgical-site infection, an accidental narcotic overdose, and an accidental incisional burn.
Among all hospital readmissions, 10/19 (53%) carried diagnoses of abdominal pain, constipation, dehydration, or oral intake intolerance without anatomic abnormality identified. These readmissions were discharged after symptomatic improvement with supportive care, including IV fluid and medication administration, and did not require invasive tests or procedures. The remaining 9/19 readmissions carried anatomic diagnoses or required therapeutic procedures (reoperation or endoscopic procedure). These diagnoses included gastric torsion, marginal ulcer, anastomotic stricture, and superior mesenteric vein thrombosis.
Discussion
Patients with lower level of education (≤12th grade) were greater than three times more likely to visit the ED or be readmitted postoperatively following bariatric surgery at our center, compared with those with a higher level of education. A very small proportion of our patients (7.5%) had a less than perfect score on the REALM-SF assessment, limiting its effectiveness as a predictor of postoperative hospital visits. The majority of postoperative ED visits and readmissions carried no clearly defined anatomic diagnosis and symptom improvement occurred with supportive care, including IV fluid administration, antiemetics, and bowel regimen.
Strengths of this study include its prospective model and clearly defined objectives. This study is the first publication to prospectively show a correlation between lower education levels with postoperative hospital visits in bariatric surgery. Limitations include the single-center study population, which may limit its external validity as well as the relatively small enrollment compared with larger database studies. ED visits and readmissions occurring at outside hospitals may have not been recorded in our data if they were not reflected in our center's EMR.
Similar to the findings of prior studies, many of our bariatric center's postoperative ED visits and readmissions are preventable. Patients frequently present to the ED with complaints of nausea, dehydration, or abdominal pain, but are not found to have significant laboratory or imaging abnormalities. On admission, documented oral intake is within the recommended volume provided in the education materials, and additional reassurance is provided for symptomatic complaints within the normal realm of expectations postoperatively. Nursing administration of recommended diet further corroborates normal expectations. Despite extensive preoperative and postoperative counseling in addition to written information providing reinforcement of diet and milestones, many of these admissions are felt to be due to dietary noncompliance or poor understanding of the early satiety caused by bariatric procedures. These factors are significant not only in the sense of readmission rates and hospital costs but also can translate to patients' overall weight-loss success if they are unable to execute the appropriate postoperative recommendations.
Patients most likely to have difficulty understanding postoperative instructions are at higher risk for increased healthcare utilization postoperatively leading to increased costs and likely patient dissatisfaction. This presents an opportunity to improve care. If lower education level is associated with increased rate of readmission, we may be able to improve outcomes through personalized attention and optimization of preoperative education and materials. Low education level can be easily identified preoperatively and these patients may benefit from extra education sessions preoperatively as well increased efforts at postoperative monitoring. Potential intervention includes scheduled calls to check in on these patients postoperatively. This creates an additional opportunity to provide counseling or encourage early clinic follow-up rather than more costly after-hours ED visits. If increased vigilance in preoperative education and postoperative monitoring of these patients can decrease postoperative healthcare utilization, we may be able to significantly decrease bariatric surgery costs.
Footnotes
Disclosure Statement
The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.