A Prospective Study on the Diagnoses for Abdominal Pain After Bariatric Surgery: The OPERATE Study

The OPERATE (abdOminal Pain aftER bAriaTric surgEry) study is a prospective cohort study to evaluate multiple aspects of abdominal pain in patients after BS, at the Spaarne Gasthuis Hospital Hoofddorp, a dutch bariatric center. At our hospital, we have our own multidisciplinary team and 24/7 on call bariatric service. We perform 800 surgeries annually, with four bariatric surgeons, each with over 12 years of experience. Inclusion started in December 2020 and is ongoing. The entire episodes of all patients with abdominal pain after BS visiting the outpatient clinic or emergency department of the Spaarne Gasthuis are included. Figure 1 depicts examples of an episode. An episode starts during the first visit with abdominal pain within the study period and is completed when a patient did not visit the hospital because of abdominal pain for three consecutive months. After these 3 months, a new episode is initiated if a patient returns to the hospital with abdominal pain. All patients are followed throughout the entire episode of abdominal pain, which includes the first and all follow-up visits, peer consultations, diagnostic tests, readmissions, and reoperations for the pain. Enrolled patients receive regular treatment, and the study does not affect the patient’s visits, the choice of additional tests, follow-up, or treatment. At the end of each episode, it is determined whether the symptoms remained or whether the abdominal pain resolved since the last (follow-up) visit. Patients treated for abdominal pain in another hospital at the same time or with a BS procedure other than (revision) laparoscopic Roux-en-Y gastric bypass, laparoscopic sleeve gastrectomy, or laparoscopic one anastomosis gastric bypass are excluded from the study. The study protocol was approved by the Local Ethics Committee. The principles of the Declaration of Helsinki are followed.

The current substudy included patients who visited the hospital for abdominal pain between December 2020 and December 2021. Only the first episode of each patient with abdominal pain within this study period was included. In conclusion, in this study, each patient represents one episode. Ongoing episodes after November 1st, 2022, were excluded.

At the end of each visit and/or episode, it was assessed whether a patient met the criteria for one or more diagnoses. A distinction was made between presumed diagnoses and definitive diagnoses (Table 1). Two bariatric surgeons together with two researchers have jointly established the criteria for the most common diagnoses for abdominal pain, which was based on the golden standard. For example, (1) an internal herniation was called a presumed diagnosis if it was present on an abdominal CT-scan and a definitive diagnosis if it was confirmed during surgery and (2) an anastomotic ulcer was called a presumed diagnosis if symptoms improved during follow-up after treatment initiation and a definitive diagnosis after confirmation with a gastroscopy. When the criteria for a presumed diagnosis or definitive diagnosis were not met at the end of the episode, the complaints were classified as unexplained complaints. Because the official criteria for IBS and constipation, such as the ROME criteria, could not be derived from the clinical data, these diagnoses were only classified as a presumed diagnosis.

Primary outcome measures are the number of presumed diagnoses, definitive diagnoses, and unexplained complaints for abdominal pain in patients after BS during the first visit and at the end of an episode. Secondary outcomes are (1) the number of follow-up visits and their yield in the diagnostic process, (2) the number and duration of readmissions and number of reoperations, (3) the time it takes to establish a presumed diagnosis or definitive diagnosis, (4) the timeline between BS and the various diagnoses of abdominal pain, and finally, (5) whether complaints have resolved at the end of an episode.

Baseline characteristics were collected from electronic patient records. Emergency department visits, readmission, and reoperations for abdominal pain after BS, before the start of the study period, were recorded. Before the start of the study, standardized reporting by the radiologist or gastroenterologist was implemented for abdominal CT-scans, ultrasounds, gastroscopies, barium swallow X-rays, and abdominal X-rays. Bariatric surgeons also completed a standardized surgery report in case of a reoperation. All data was stored in an electronic case record form in a web-based data capture-system (Castor EDC). The database was anonymized by a study number. Three researchers (NO, MV, and SV) having exclusive access to the database gathered the data and are able to re-identify patients based on their study number.

Data were reported as mean and standard deviation (SD) for normally distributed variables and as median and interquartile range (IQR) for variables with a skewed distribution. Continuous variables between groups were compared using the independent sample t-test, the non-parametric Mann-Whitney U test, or Kruskal Wallis test, depending on their distribution. The chi-square test was used for categorical variables. Logistic regression analysis was used to evaluate associations between patient characteristics and diagnoses. Variables with a p value of 0.10 or less in the univariable analysis were selected for the multivariable model. Stepwise backward elimination was used to derive a final model: in each subsequent step, the least significant variable in the model was removed until all remaining variables had individual p values smaller than 0.10. All p values are two-sided, and p values below 0.05 were considered statistically significant. Data were analyzed using Statistical Package for Social Sciences (SPSS) version 24 for windows (IBM Corporation, New York, USA). Tables and figures were made using Microsoft and SPSS.

From December 2020 until December 2021, 448 patients after BS visited the hospital for abdominal pain. Seven patients were excluded: three patients with incomplete episodes, two patients who were simultaneously treated in another hospital, and two patients with unclear anatomy of the gastric bypass. Baseline characteristics of the remaining 441 patients are presented in Table 2. The BS was performed in our hospital in 363 (87.7%) patients. Baseline characteristics of patients with their first visit at the outpatient clinic (204 (46.2%)) or emergency department (237 (53.7%)) were compared. A significant and clinically relevant difference was found only for the time after surgery in months, with 44.0 (IQR 16.5–73.0) and 26.0 (IQR 6.0–59.0) months for outpatient clinic and emergency department visits, respectively (p < 0.001).

At the end of the episode, 88 (20.0%) patients had a presumed diagnosis, 183 (41.5%) patients a definitive diagnosis, and 170 (38.5%) patients had unexplained complaints. Twenty-six patients had more than one diagnosis at the end of the episode. In 12 patients, the second or third diagnosis was a complication of the first diagnosis, for instance, bile leakage after cholecystectomy. The remaining patients had two or three unrelated diagnoses during their episode.

Internal herniation and cholecystolithiasis were the most common presumed diagnoses during first visit (Fig. 2). The most common diagnoses at the end of follow-up were cholecystolithiasis in 60 (13.6%), ulcers in 43 (9.8%), and internal herniation in 33 (7.5%) patients.

A total of 136 patients (30.8%) had only one follow-up visit, whereas the majority had two to five follow-up visits in descending order (Fig. 3a). Unlike the first visit, most follow-up visits took place at the outpatient clinic (Fig. 3b). In addition, 77 (17.5%) patients had 1–3 peer consultations: gastroenterology (49.5%), general internal medicine (20.0%), pain management team (12.6%), urology (10.5%), gynecology (5.3%), and plastic surgery (2.1%). There were 260 readmissions across 212 patients, ranging from 1 to 4 per patient. Of these, 77 (29.6%) readmissions were for elective surgery. Readmissions lasted for 1 day in 120 (46.2%) cases and 51 (19.6%) lasted four or more days. In total, 186 reoperations were performed in 164 patients, with 1–3 reoperations per patient. Diagnostic laparoscopy was the most performed surgery followed by laparoscopic cholecystectomy (89 [47.8%] and 57 [30.6%], respectively). Data concerning readmissions and reoperations is shown in Supplemental Table S1.

Figure 3c shows that the number of presumed diagnoses established in patients with a certain amount of follow-up visits decreased, whereas the number of definitive diagnoses almost remained the same in patients with 2–5 follow-up visits. The percentage of patients with unexplained complaints decreased over the number of follow-up visits, whereas the proportion of definitive diagnoses increased. Despite the fact that patients often had more than 3 follow-up visits, Fig. 3d shows that a diagnosis was frequently made within the first visit or first three follow-up visits.

Median (IQR) time to a presumed diagnosis, definitive diagnosis, or unexplained complaints from the start of the episode was 16.0 (3.8–44.5), 2.0 (0.0–31.5), and 13.5 (1.0–53.8) days, respectively (p < 0.001). Internal herniation, cholecystolithiasis, and appendicitis were usually diagnosed during the first visit or first follow-up visit. Anterior cutaneous nerve entrapment syndrome (ACNES), abdominal wall herniation, IBS, and constipation were usually diagnosed later on in the episode (Fig. 4a). These findings are consistent with Fig. 4b. Figure 4c demonstrates that there were no notable differences for specific diagnoses in time after BS. This was also not the case for the time after BS and a presumed diagnosis, definitive diagnosis, or unexplained complaints (p = 0.929).

The multivariable model on the association between patient characteristics and diagnosis (presumed diagnoses/definitive diagnoses or unexplained complaints) revealed that patients with IBS more often had unexplained complaints at the end of the episode compared to patients without IBS (OR 95%CI: 4.46 [1.46–13.65], p = 0.009. The association between presence of a history of psychiatric diseases and unexplained complaints did not reach statistical significance (OR 95%CI: 1.63 [0.98–2.73]), p = 0.062). Patients with IBS had a significant higher probability of unexplained complaints when compared to patients with a definitive diagnosis (OR 95%CI: 4.27 [1.39–13.15], p = 0.011). A lower BMI at first presentation, IBS, and previous abdominal surgery were associated with a presumed diagnosis if compared to patients with a definitive diagnosis (OR 95%CI: 1.07 [1.01–1.13], p = 0.013; 6.74 [1.77–25.70], p = 0.005; and 1.45 [1.08–1.96], p = 0.014, respectively).

At the end of the episode, 71 patients (16.1%) still experienced abdominal pain of which 29 (40.1%) had unexplained complaints. In 96 (21.7%) patients, it was not clear whether complaints had been resolved due to lack of follow-up or wait-and-see policy. Patients with IBS were most likely to have persistent complaints and in 5 (8.3%) patients treated for cholecystolithiasis complaints persisted without an alternative diagnosis. Data is shown in Supplemental Tables S2 and S3.

All previous visits, readmissions, reoperations, and diagnoses for abdominal pain before the start of the study are not included in the analyses. This makes the study less comparable to current retrospective cohort studies. On the other hand, this design results in a larger group of patients, and it better reflects daily practice in the outpatient clinic and emergency department. In patients with previous unexplained complaints, more acute causes were possible already ruled out during earlier episodes. Therefore, the diagnostic process could have focused more on another set of diagnoses than in patients with a first presentation for abdominal pain. Future research on sequential episodes in the OPERATE study will provide more information on the impact of these patients. This study was started in the COVID pandemic. This could have impacted the number of visits and the duration of the episodes. The numbers of several diagnoses were too small to evaluate in what time after surgery they occurred. Furthermore, although no associations were observed between type of operation and outcomes, Roux-en-Y gastric bypass is still the number one of BS performed in the Netherlands, and there are too little sleeve gastrectomies and one-anastomosis gastric bypass surgeries to compare with. This is important because current literature shows that the numbers of emergency department visits and readmissions for abdominal pain after sleeve gastrectomy are quite similar to those of the Roux-en-Y gastric bypass [33]. Lastly, socio-economic factors are not included in this study.