Nationwide Outcomes and Readmission After Pediatric Laparoscopic and Open Fundoplication

Introduction

Gastroesophageal reflux is a common and normal phenomenon in infants, characterized by regurgitation of stomach contents into the esophagus due to transient relaxation of the lower esophageal sphincter. Reflux typically improves in infancy, with 5% experiencing persistent symptoms at the end of the first year of life.^1,2 In older children, reflux symptoms have an increasing incidence with age. ³ A diagnosis of gastrointestinal reflux disease (GERD) may be made with persistent symptoms or complications such as growth failure, frequent regurgitation, sleep disturbances, feeding difficulty, or respiratory compromise.^4,5

Patients with GERD are initially managed with dietary and lifestyle modifications, including postural changes and dietary alterations. Acid-suppressive therapy is used for persistent symptoms with good response^6–8 ; however, there are risks associated with long-term use of antisecretory agents.^9–11 Surgical intervention is typically reserved for cases refractory to pharmacotherapy. ¹² Although indications for surgical intervention are somewhat unclear ¹³ and without high-quality evidence, ¹² common indications include airway disease, neurologic impairment, recurrent aspiration, and complex or frequent feeding requirements. ¹⁴ However, current guidelines do not define failure of medical therapy or outline specific recommendations such as timing or optimal patient selection.

The most commonly performed surgical procedure for GERD is the Nissen fundoplication. Key steps of the procedure involve division of the short gastric arteries, developing a retroesophageal window, limiting esophageal mobilization, ensuring adequate intraabdominal esophageal length, and wrapping the fundus of the stomach 360° posteriorly around the distal esophagus. ¹⁵ Although this has historically been performed via an upper midline laparotomy, proponents of the laparoscopic technique believe this to be a safer approach, and have utilized it since the 1990s. Retrospective studies have found decreased perioperative morbidity ¹⁶ and complications ¹⁷ in patients who underwent laparoscopic fundoplication. Financially, the laparoscopic procedure has been associated with lower lengths of stay (LOS), translating to reduced cost. ¹⁸ However, some studies have called into question the superiority of laparoscopy compared to open fundoplication with regard to short- and long-term clinical outcomes. This study seeks to explore index admission and readmission outcomes comparing laparoscopic versus open operative techniques from a nationally representative sample in an attempt to clarify this discrepancy and provide clearer recommendations on the safety of each procedure. We hypothesize that laparoscopic fundoplication will have superior outcomes to the open procedure.

Methods

The Healthcare Cost and Utilization Project (HCUP) is the largest Federal-State-Industry partnership providing national encounter-level health care data in the United States. HCUP provides researchers with a family of databases, including the Nationwide Readmissions Database (NRD). ¹⁹ The NRD addresses a large gap in health care data – the lack of nationally representative information on hospital readmissions for all ages. This database is released annually and comprises 15 million admissions across 22 states, representing 51% of all U.S. hospitalizations. The NRD provides the unique ability of tracking patients across hospitalizations at different hospitals. It contains multistate data from ∼17 million discharges each year and is weighted to estimate roughly 36 million discharges nationally. NRD data include deidentified patient demographics, diagnoses, procedure codes, length of stay, and characteristics of the treating hospital for each patient encounter. For transfers or same-day discharge and admission between different hospitals, these multiple records are collapsed into one by the NRD to represent a single hospitalization. The NRD is extensively processed to ensure that readmission counts are accurate, adding validity to the use of these data. ¹⁹ The NRD data are also extensively processed to ensure that readmissions are counted as accurately as possible. The NRD greatly reduces the likelihood of this type of miscoding by collapsing multiple records into one if they involve a transfer or same-day event such as discharge and admission between different hospitals. As the NRD is based on annual data from individual State Inpatient Databases (SID), it cannot track readmissions that cross calendar years or occur in a different state. However, the NRD is still the most comprehensive and accurate source of U.S. hospital readmissions data, making it the ideal database for studies on nationwide readmissions.

From 2010 to 2014, the NRD was queried to identify patients 0 to 18 years old who underwent a fundoplication using the International Classification of Diseases, Ninth Revision Clinical Modification (ICD-9CM) procedure code for either open (44.66) or laparoscopic (44.67) fundoplication. Patient demographics, hospital factors, and outcomes were analyzed. Hospital readmissions were defined as admission within 30 days and within the year of index discharge (overall readmission). Elective and nonelective readmissions were included in analysis. ICD-9CM diagnosis and procedure codes are provided by the NRD for index hospitalization and readmissions. Fundoplication complications were divided into two categories: major and minor. Major complications were defined as bowel obstruction, gastrostomy closure/resiting, gastrostomy revision to gastrojejunostomy, jejunostomy, or ventriculoperitoneal (VP) shunt revision. Minor complications were defined as dysphagia, gas bloat, gastric functional disorder, or vomiting. Associated comorbidities, operative procedures, and complications were examined.

IBM SPSS Statistics version 26 (International Business Machines Corp., Armonk, NY) was used for statistical analysis. Demographics are reported as n (%). Quantitative variables are reported as mean ± standard deviation for normally distributed variables or median [interquartile range (IQR)] for nonparametric variables, and analyzed using Student's t-test or Mann–Whitney U test, as appropriate. Propensity score (PS)-matched analysis was performed by matching those undergoing laparoscopic fundoplication with patients receiving open fundoplication based on 87 covariates, including gender, age, Charlson Comorbidities, associated diagnoses, congenital anomalies, and perinatal conditions. Chi-squared analysis was used to compare categorical variables between cohorts. Statistical significance was defined as P < .05. Results were weighted for national estimates as per HCUP utilizations guideline. ¹⁹ This retrospective comparative analysis was deemed exempt from review by (Univeristy of Miami Hospital System) Institutional Review Board.

Results

Discussion

While many studies have examined long- and short-term outcomes of fundoplication in the pediatric population, this is one of the first studies to explore nationwide outcomes and readmissions comparing laparoscopic and open fundoplication. We have found that laparoscopic procedures are performed more often than open. In PS matched analysis, laparoscopic fundoplication was associated with lower short-term readmission rates and fewer major and minor complications than open fundoplication. In addition, open fundoplication was associated with higher rates of pneumonia and infections postoperatively. These differences remained when those undergoing concomitant gastrostomy were excluded.

Although studies have confirmed safety and improved outcomes with laparoscopic fundoplication,^15,16 there is some discrepancy in the literature.^{16–18,20,21} A few prospective randomized control trials (RCTs) have called into question the superiority of laparoscopy compared to open fundoplication. In 2020, Fyhn found no short-term difference in open versus laparoscopic outcomes in the pediatric literature. ²² A trial of 39 children found no short-term postoperative differences in dysphagia, recovery, recurrence or redo fundoplication between the open and laparoscopic groups. ²⁰ A study that evaluated a RCT at 4 years found no difference in control of reflux or quality of life. ²³ Similarly, Papandria et al. found no differences in short- or long-term clinical outcomes between the two techniques in children younger than 2 years of age and noted higher cost and longer operating time in the laparoscopic group. ²³ A 2015 RCT demonstrated a higher recurrence rate of GERD symptoms in the laparoscopic group compared to the open technique group. ²⁴ To contrast, a study performed in Japan in 2019 found higher rates of postoperative bowel obstruction in open procedures. ²⁵ Clearly, the literature is discrepant with regard to the superiority of a particular technique in pediatric fundoplication.

Recent trials have also examined the cost effectiveness and long-term results for reducing GERD symptoms. When considering cost, Papandria et al. concluded that laparoscopy led to higher operative costs with similar outcomes. ²³ Dekonenko notes decreased retching rate, length of stay, and morbidity with laparoscopy despite the higher cost and longer operating room (OR) time. ²⁶ Our analysis, using PS matching, found a significant decrease in hospital costs in the laparoscopic cohort, likely due to a shorter hospital LOS. With regard to symptomatic relief, two studies described no change in symptoms such as dysphagia, regurgitation, or need for redo fundoplication based on surgical approach.^17,18 In our nationwide study, we found higher rates of minor (dysphagia, gas bloat, gastric dysfunction, and vomiting) and major (gastrojejunostomy/jejunostomy tube, bowel obstruction, gastric fistula closure/resiting, and VP shunt replacement) complications in the open group. No patient required redo Nissen fundoplication within the yearlong study period. Our analysis demonstrates that laparoscopic fundoplication is associated with similar and possibly superior symptomatic control and need for fewer interventions.

There is significant complexity regarding the diagnosis of GERD, indications for surgery, and heterogeneity of populations undergoing fundoplication. For this reason, the American Pediatric Surgical Association (APSA) Outcomes and Evidence Based Practice committee performed a review of surgical therapies focusing on the effectiveness of open versus laparoscopic fundoplication, including two RCTs and one systematic review.^20,27,28 Neither RCT found a difference with regard to early postoperative outcomes. However, they cautioned interpretation of the results due to low sample sizes, issues with study power, and study design. They ultimately concluded that laparoscopic fundoplication is most likely comparable to open fundoplication for short-term clinical outcomes and long-term data are lacking.

Another consideration is the decision to operate at all. As mentioned, fundoplication is not the first line treatment in GERD. Only 5% of infant patients will continue to have GERD symptoms by the time they reach 1 year of age.^1,2 However, there are concerns for risk of antisecretory agents after long-term use. Adverse effects have been reported in at least 23% of patients treated with H2 blockers and 34% of those treated with proton pump inhibitors. Most of these include idiosyncratic reactions (headache, nausea, diarrhea, constipation). Other less common risks include drug-drug interactions (phenytoin, diazepam, warfarin, clopidogrel), rebound hypersecretion after proton pump inhibitor (PPI) discontinuation, osteoporosis, hypomagnesemia and increased rates of pneumonia, gastrointestinal infections, and clostridium difficile-associated diarrhea. ²⁹ In addition, recent long-term follow-up studies after fundoplication have found that after 5 years, the number of symptom-free patients decline. ³⁰ The decision to move forward with fundoplication should be made on a case-by-case basis in the pediatric population, weighing the risk and benefits of long-term medications versus operative risks. As of yet, no prospective trials have investigated the outcomes of long-term medical therapy versus operative intervention.

A possible confounding aspect to the data is the surgeon's views on the technical aspects and challenges of performing a laparoscopic fundoplication in newborns, especially low-weight infants. Historically, surgeons would not perform laparoscopy on children less than 5 kg. However, multiple studies have demonstrated safety and efficacy in neonates even less than 3 kg.^31–34 It is possible that surgeons more facile with laparoscopy account for the superior outcomes with this technique. A study by Meehan studying the learning curve of laparoscopic fundoplication in infants showed that as surgeons had more experience, their rate of conversion and length of operation decreased. ³² The data may be biased toward surgeons who do the cases more often electing to do them laparoscopically versus surgeons who do this procedure less frequently electing to operate open. There is a growing concern in the pediatric surgery community of the lack of experience this current generation of pediatric surgeons will have with fundoplications. The rate of fundoplication in children has decreased threefold in the last 10 years, while fellow trainee volume has decreased twofold in the same time period.^35,36 There may be a need for changes in the curriculum to aid current pediatric surgery trainees with fundoplication experience in the future.

This study has limitations due to the retrospective nature of the database, and limitations of using national databases. First, the study design is a retrospective review of the NRD, which lends itself to errors in data sampling, collection measures, usage of the ICD-9 coding scheme, and possible administrative errors during data entry. Important perioperative factors such as operative time, concurrent medical therapies, indication for surgery, and timing of index admission complications are not recorded in NRD, and therefore are not included in this analysis. Large databases lack granularity in diagnosis, as they rely on ICD-9 diagnosis. The lack of timelines could lead to measurement bias, as the timing of complications during the index hospitalization compared to operative dates cannot be confirmed. The NRD does not follow patients across years or across state lines if readmitted. Although HCUP estimates that less than 5% of readmissions occur across state lines, the inability to track patients across years may cause an underestimation of readmission outcomes, including fundoplication complications in the long term.

Conclusion

The majority of fundoplications are being performed in newborns and are being done laparoscopically, which are associated with lower complications and postoperative readmission rates compared to open fundoplications.