- Email: [email protected]
Exploring regional variability in utilization of antireflux surgery in children
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Association for Academic Surgery
Exploring regional variability in utilization of antireflux surgery in children Heather L. Short, MD,a Wanzhe Zhu, BS,b Courtney McCracken, PhD,c Curtis Travers, MPH,c Lance A. Waller, PhD,b and Mehul V. Raval, MD, MSa,* a
Division of Pediatric Surgery, Department of Surgery, Children’s Healthcare of Atlanta, Emory University School of Medicine, Atlanta, Georgia b Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University School of Medicine, Atlanta, Georgia c Department of Pediatrics, Emory University School of Medicine, Atlanta, Georgia
article info
abstract
Article history:
Background: There is significant variation surrounding the indications, surgical approaches,
Received 2 December 2016
and outcomes for children undergoing antireflux procedures (ARPs) resulting in geographic
Received in revised form
variation of care. Our purpose was to quantify this geographic variation in the utilization of
12 January 2017
ARPs in children.
Accepted 28 February 2017
Methods: A cross-sectional analysis of the 2009 Kid’s Inpatient Database was performed to
Available online 8 March 2017
identify patients with gastroesophageal reflux disease or associated diagnoses. Regional surgical utilization rates were determined, and a mixed effects model was used to identify
Keywords:
factors associated with the use of ARPs.
Pediatric surgery
Results: Of the 148,959 patients with a diagnosis of interest, 4848 (3.3%) underwent an ARP
Antireflux surgery
with 2376 (49%) undergoing a laparoscopic procedure. The Northeast (2.0%) and Midwest
Variation in care
(2.2%) had the lowest overall utilization of surgery, compared with the South (3.3%) and
GERD
West (3.4%). After adjustment for age, case-mix, and surgical approach, variation persisted
Regional variation
with the West and the South demonstrating almost two times the odds of undergoing an ARP compared with the Northeast. Surgical utilization rates are independent of state-level volume with some of the highest case volume states having surgical utilization rates below the national rate. In the West, the use of laparoscopy correlated with overall utilization of surgery, whereas surgical approach was not correlated with ARP use in the South. Conclusions: Significant regional variation in ARP utilization exists that cannot be explained entirely by differences in patient age, race/ethnicity, case-mix, and surgical approach. In order to decrease variation in care, further research is warranted to establish consensus guidelines regarding indications for the use ARPs for children. ª 2017 Elsevier Inc. All rights reserved.
Institutional Review Board approval: Exempt. * Corresponding author. Division of Pediatric Surgery, Department of Surgery, Children’s Healthcare of Atlanta, Emory University School of Medicine, 1405 Clifton Road NE, Atlanta, GA 30322. Tel.: þ1 404-785-0781; fax: þ1 404 785-0800. E-mail address: [email protected] (M.V. Raval). 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.02.075
50
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
Introduction Gastroesophageal reflux is the normal physiological passage of gastric contents into the esophagus that can occur several times a day in infants and children.1 Most of these cases resolve spontaneously by 18 mo of life and do not cause any symptoms or complications.2 However, physiological reflux can progress to gastroesophageal reflux disease (GERD) and result in numerous pathologies that can negatively impact a child’s well-being. Medical management, specifically the use of proton pump inhibitors, is the mainstay of treatment for symptomatic reflux; however, antireflux procedures (ARPs) remain an option for children whose symptoms do not respond to medical management. Although there appear to be obvious benefits to the surgical treatment of GERD, there is significant variation in the
indications, preoperative evaluation, surgical approach, and outcomes for children undergoing ARPs.3-5 Nusrat and Bielefeldt demonstrated a fivefold variation in regional ARP utilization rates in the United States for the treatment of GERD in adults.6 In addition, there was a marked increase in the use of ARPs for treatment of GERD in adults with the widespread introduction of laparoscopy in the 1990s.7-9 National surgical utilization rates declined steadily from 1999 to 200310; however, the volume of elective adult ARPs has remained relatively stable since about 2005.11 Surprisingly, when a pediatric population was examined between 1996 and 2003, no similar trends were identified.12 The purpose of this study was to examine contemporary regional and state-level variability in utilization of ARPs to treat children with GERD and its complications. We hypothesized that regional and state level rates of surgical utilization would correlate with use of laparoscopy.
Table 1 e Demographics (age, race, and comorbidities), including all diagnoses with proportion that underwent ARPs. All patients
Total N Mean number of total diagnoses (SD)y
148,969
*
Antireflux surgery Yes
No
4230 (2.8%)
144,739 (97.2%)
7.672 (4.82)
9.752 (6.284)
7.611 (4.759)
71,089 (47.7%)
1939 (45.8%)
69,150 (47.8%)
69,938 (46.9%)
2009 (47.5%)
67,929 (46.9%)
Between 1 and 10 y
39,147 (26.3%)
1565 (37.0%)
37,582 (26.0%)
>10 y
39,468 (26.5%)
636 (15.0%)
38,832 (26.8%)
White
71,025 (47.7%)
2065 (48.8%)
68,960 (47.6%)
Black
21,201 (14.2%)
624 (14.8%)
20,577 (14.2%)
Female, N (%)z Age, N (%)x <1 y
Race, N (%)k
Hispanic
23,623 (15.9%)
668 (15.8%)
22,955 (15.9%)
Other
10,503 (7.1%)
332 (7.8%)
10,171 (7.0%)
26,028 (17.5%)
508 (12.0%)
25,520 (17.6%)
Region, N (%) Northeast Midwest
34,800 (23.4%)
763 (18.0%)
34,037 (23.5%)
South
56,457 (37.9%)
1884 (44.5%)
54,573 (37.7%)
West
31,684 (21.3%)
1075 (25.4%)
30,609 (21.1%)
4300 (2.9%)
184 (4.3%)
4116 (2.8%)
Contributing diagnoses (more than one may apply), N (%) Esophagitis, N (%) Esophageal stricture, N (%)
11,793 (7.9%)
Aspiration, N (%) Apnea, N (%) Failure to thrive, N (%) Barrett’s esophagus, N (%) Paraesophageal hernia/hiatal hernia, N (%) Esophageal reflux, N (%) Esophageal ulcer, N (%) *
828 (0.6%)
Dysphagia, N (%)
54 (1.3%)
774 (0.5%)
553 (13.1%)
11,240 (7.8%)
10,994 (7.4%)
414 (9.8%)
10,580 (7.3%)
8987 (6.0%)
143 (3.4%)
8844 (6.1%)
27,156 (18.2%)
1366 (32.3%)
25,790 (17.8%)
123 (0.1%)
23 (0.5%)
100 (0.1%)
2820 (1.9%)
597 (14.1%)
2223 (1.5%)
105,265 (70.7%) 445 (0.3%)
3840 (90.8%)
101,425 (70.1%)
13 (0.3%)
Of the 148,969 patients, 22,904 patients (15.4%) are missing either age or race. Number of diagnoses range from 1 to 35 and may contain other diagnoses not associated with ARP. z Gender is missing for 278 patients (0.2%). x Age is missing for 416 patients (0.3%). k Race is missing for 22,617 patients (15.2%), this includes all patients from Ohio, North Carolina, West Virginia, and Minnesota. y
432 (0.3%)
51
s h o r t e t a l v a r i a b i l i t y i n a n t i r e fl u x s u r g e r y u s e
Methods A cross-sectional analysis of the Agency for Healthcare Research and Quality (AHRQ) sponsored Healthcare Cost and Utilization Project (HCUP) Kids’ Inpatient Database (KID) was performed to identify children with a reflux associated diagnosis in 2009. The KID is a national administrative dataset containing hospital discharge-level data for patients aged <21 y. The dataset contains data from 44 states and more than 4100 US hospitals including information for over 3 million pediatric discharges per year sampled from all community, nonrehabilitation hospitals of states participating in HCUP. National estimates of discharges were generated using discharge weights provided by HCUP that are based on the poststratification of hospitals according to characteristics including ownership, bed size, teaching status, rural/urban location, region of the United States, and being a freestanding children’s hospital. Records from each hospital are randomly sampled to include 10% of uncomplicated births and 80% of complicated births because less variation in outcomes among uncomplicated births is expected. More than 100 clinical and nonclinical data elements are available for each discharge record including patient demographics, hospital charges, and diagnosis and procedure codes.13 Using the International Classification of Diseases, Ninth Revision, Clinical Manifestation, all patients (aged <21 years) with a diagnosis of GERD (530.81) or one of its related diagnoses, including esophagitis (530.10, 530.11), esophageal stricture (530.3), dysphagia (787.20, 787.21, 787.22, 787.23, 787.24), aspiration (507.0), apnea (786.03), failure to thrive (783.41, 779.34), Barrett’s esophagus (530.85), esophageal ulcer (530.20), or hiatal hernia (553.3) were identified. Children who underwent an ARP were identified with International Classification of Diseases, Ninth Revision, procedure codes for “other procedures for creation of esophagogastric sphincteric competence” (44.66) or “laparoscopic procedures for creation of esophagogastric sphincteric competence” (44.67). Demographics obtained from the database included age, race, number of diagnoses, location of hospital visit (hospital ID and state). Hospital-related information obtained from the database included hospital ID, state, and census region (Northeast, South, Midwest, and West). Individual state and regional surgical utilization rates were calculated by tabulating the number of ARPs performed for a given state or region, then dividing by the total number of records with at least one of the diagnoses identified above for that state. Descriptive statistics were calculated for all variables of interest and include means and standard deviations or counts and percentages, as appropriate. A mixed effects model was used to determine which regions and individual states were high utilizers of surgery and to identify patient and hospital factors associated with open versus laparoscopic procedures. A separate mixed effects logistic model was used to identify patient and hospital factors associated with the decision to perform ARPs on patients diagnosed with GERD or one of its related diagnoses, with the hospital’s state being the random effect. The addition of a random intercept improved model fit considerably over a fixed effect model as measured by
Table 2 e Factors associated with utilization of antireflux surgery based on a multivariate mixed-effects logistic model. OR (95% CI Lower limit-Upper limit)
P
Age <1 y
1.00
Reference
Between 1 and 10 y
1.23 (1.15-1.31)
<0.001
>10 y
0.48 (0.44-0.53)
<0.001
White
1.00
Reference
Black
1.01 (0.93-1.09)
0.900
Hispanic
0.87 (0.80-0.94)
<0.001
Race
Other Number of diagnoses*
1.02 (0.92-1.14)
0.686
1.07 (1.06-1.07)
<0.001
1.00
Reference
Region Northeast Midwest
1.30 (0.81-2.08)
0.271
South
1.97 (1.27-3.03)
0.002
West
1.85 (1.18-2.89)
0.007
Esophagitis
2.83 (2.39-3.34)
<0.001
Esophageal stricture
2.40 (1.75-3.30)
<0.001
Dysphagia
2.51 (2.29-2.76)
<0.001
Aspiration
2.65 (2.39-2.95)
<0.001
Apnea
1.20 (1.03-1.40)
0.022
Failure to thrive
4.27 (4.00-4.56)
<0.001
Barrett’s esophagus
9.20 (5.43-15.60)
<0.001
Paraesophageal/ hiatal hernia
22.23 (19.88-24.87)
<0.001
Esophageal reflux
10.93 (9.90-12.06)
<0.001
Esophageal ulcer
0.47 (0.24-0.91)
Contributing diagnoses
*
0.026
Number of total diagnoses ranges from 1 to 35.
Akaike’s Information Criterion. Results from the mixed effect models are presented as odds ratios with associated 95% confidence intervals (CIs). Analyses were performed in R 3.2.2 using the lme4 package function for mixed effects models.14,15 A P < 0.05 was considered statistically significant.
Results Of the 148,969 patients with one or more of the diagnoses of interest, 4230 (2.84%; 95% CI 2.76-2.92) patients underwent an ARP. Of the 4230 patients who underwent an ARP, 2095 (49.53%) underwent a laparoscopic procedure, 2125 (50.24%) underwent an open procedure, and 10 (0.23%) were indeterminate. Demographics of the cohort are presented in Table 1. Hispanics were less likely to undergo ARP compared with non-Hispanic whites (OR 0.87; 95% CI 0.80-0.94). GERD was the most common indication (91%) for ARPs (Table 1).
52
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
Fig. 1 e Panel 1. State-specific ratios of laparoscopic utilization rates over ARP utilization rates from the 2009 KID. Panel 2. Choropleth map of the United States demonstrating state-by-state laparoscopy utilization rates. States shaded in gray are not reported in the 2009 KID. (Color version of the figure is available online.)
Using the population aged <1 y as a reference, children aged between 1 and 10 y had increased odds of undergoing an ARP (OR 1.23; 95% CI 1.15-1.31). Older children (aged 11-20 y), however, had lower ARP utilization compared with children aged <1 y (OR 0.48; 95% CI 0.44-0.53). The number of diagnoses per patient also had a positive impact on ARP utilization. Each additional diagnosis increases the odds of a patient undergoing ARP by 7% (OR 1.07; 95% CI 1.06-1.07; Table 2). The Northeast (2.0%) and Midwest (2.2%) had the lowest overall utilization of surgery, compared with the South (3.3%) and West (3.4%). After adjustment for age, case-mix, and surgical approach and allowing for state specific utilization rates, regional variation persisted with the West and the South demonstrating close to two times the odds of undergoing an ARP compared with the Northeast (Table 2). Surgical utilization rates appeared to be independent of state-level case volume with some of the highest case volume states (California, New York, and Ohio) having surgical utilization
rates below the national rate (Fig. 1, Panel 1). There appears to be an overall bias toward ARP utilization in the West and South, whereas there is more state-to-state variability in the Northeast and Midwest (Fig. 1, Panel 2). Each of the diagnoses of interest was statistically significantly predictors of a patient undergoing an ARP. A patient diagnosed with esophagitis, esophageal stricture, dysphagia, or aspiration had about 2.5 times the odds of undergoing an ARP compared with patients who did not have that diagnosis. A patient diagnosed with Barrett’s esophagus or esophageal reflux had close to 10 times the odds of undergoing an ARP. A patient diagnosed with hernia or hiatal hernia had the highest odds of undergoing an ARP. In contrast, a patient diagnosed with apnea or esophageal ulcer had the lowest odds of undergoing an ARP (Table 2). The effect of these diagnoses have on ARP utilization is multiplicative so a patient diagnosed with aspiration and failure to thrive had over 11 times the odds of undergoing ARP than the same patient without these two diagnoses.
s h o r t e t a l v a r i a b i l i t y i n a n t i r e fl u x s u r g e r y u s e
Table 3 e Factors associated with the decision to perform laparoscopic procedure versus open procedure. Factor
Adjusted OR (95% CI Lower limit-Upper limit)
P
1.00
Reference
Age <1 y Between 1 and 10 y
1.21 (1.06-1.38)
0.005
>10 y
1.91 (1.59-2.31)
<0.001
White
1.00
Reference
Black
0.83 (0.70-0.98)
0.025
Hispanic
0.95 (0.80-1.12)
0.514
Other
0.89 (0.72-1.10)
0.290
0.94 (0.93-0.95)
<0.001
1.00
Reference
Race
Number of total diagnoses* Region Northeast Midwest
1.71 (0.80-3.63)
0.167
South
1.15 (0.57-2.31)
0.698
West
3.12 (1.51-6.48)
0.002
Dysphagia
1.23 (1.03-1.46)
0.019
Apnea
1.66 (1.21-2.28)
0.002
Failure to thrive
1.33 (1.17-1.50)
<0.001
*
Number of total diagnoses ranges from 1 to 35.
Many of the same factors associated with ARP utilization are also associated with the use of laparoscopic versus open technique. When undergoing ARP, children aged between 1 and 10 y have slightly higher odds of undergoing a laparoscopic procedure than a <1-year-old (OR 1.21; 95% CI 1.06-1.38). Children aged 10-20 y have twice the odds of undergoing a laparoscopic procedure (OR 1.91; 95% CI 1.59-2.31; Table 3). Each additional diagnosis also reduces the odds of a patient undergoing laparoscopy (OR 0.94; 95% CI 0.93-0.95). Using the Northeast as a reference, patients receiving care in the West have the highest odds of undergoing a laparoscopic procedure (OR 3.12; 95% CI 1.51-6.48; Table 3). Of the 10 diagnoses of interest, only dysphagia, apnea, and failure to thrive are associated with laparoscopy usage (Table 3). There is a weak but nonsignificant correlation (0.29; 95% CI 0.05-0.57; P ¼ 0.089) between laparoscopy use and state-level case volume, and we found considerable state-to-state variation in laparoscopy use within each region (Fig. 2).
Discussion Although the mainstay of treatment for GERD and its associated diagnoses is medical management, ARPs remain an option for medically refractory cases in children. There is significant variation surrounding the indications, surgical approaches, and outcomes in children undergoing ARPs resulting in geographical variation in the utilization of surgery in the United States. Significant regional variation in ARP utilization exists with the West and the South demonstrating
53
the highest surgical utilization rates, and adjustment for age, case-mix, and surgical approach did not explain this variation. In addition, surgical utilization rates were found to be independent of state-level case volumes with some of the states with the highest case volume demonstrating surgical utilization rates below the national average. Although we hypothesized that the availability of laparoscopy and surgeons proficient in laparoscopic technique could be a driver for ARP utilization, we did not find a significant correlation between surgical utilization rates and the use laparoscopy. Rather, it appears that ARP utilization rates vary more on the state level as opposed to the regional level, and case volume does not appear to drive laparoscopy utilization, which varies considerably from state-to-state, as well. Regional variation in the use of surgical procedures is often attributed to differences in physician beliefs about the indications for surgery and patient preferences with differences in illness prevalence playing only a small role. This explanation can also be applied to the state-by-state variation that we saw in our results and may even be applicable to variation among individual hospitals. Practice variation exists in other areas of medicine, but the implication of variation in when to perform an operation or not to, is that some patients may be getting procedures that they do not want or need.16 Ideally, surgical treatment should provide maximal benefit to the patient with minimal risk of complications. When surgery is overutilized, or used in a patient who is already receiving maximal benefit from nonsurgical management, that patient is put at risk unnecessarily and the system is incurring unnecessary costs. Variation in care is a driver for systemic inefficiency, and reducing variation may decrease complexity of care, encourage adherence to best practices, generate cost savings, and lead to improved outcomes for some procedures.17 Recommendations from the International Pediatric Endosurgery Group state that surgery is indicated as a treatment for GERD when the diagnosis has been confirmed by objective testing such as pH monitoring, impedance monitoring, or esophagoscopy.18 Despite this, a survey of pediatric surgeons working in hospitals that are members of the Children’s Health Care Association found that only 13% of those surveyed require a pH probe study preoperatively.19 Furthermore, it is recommended that only children who fail medical treatment of their GERD be considered for an ARP; however, many children are not trialed on medical therapy before surgery, and there are no randomized trials comparing medical therapy alone with fundoplication.1 The definition of failure of medical management remains vague and often the decision to perform ARPs is based on the presence of clinical symptoms such as pain, aspiration, failure to thrive, apnea, or acute lifethreatening events in the neonate. These factors allow for wide variation in the clinical decision making in regards to the utilization of surgery in the treatment of GERD in children. Most of the existing literature on outcomes in the surgical treatment of GERD consists of retrospective case series in which the documentation of diagnostic testing and the details of previous medical management are lacking.20-22 Moreover, many of the studies that describe positive outcomes after ARPs use subjective measures of outcome to assess their efficacy rather than objective tests to confirm the resolution of
54
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
Fig. 2 e Panel 1. State-specific ratios of ARP utilization rates from the 2009 KID demonstrating that surgical utilization rates do not correlate with the volume of patients with a diagnosis of interest. Panel 2. Choropleth map of the United States demonstrating state-by-state ARP utilization rates per 1000 diagnosis of interest. States shaded in gray are not reported in the 2009 KID. (Color version of the figure is available online.)
symptoms.21,23 This makes it difficult to come to a meaningful conclusion regarding postoperative outcomes in children undergoing ARPs. The deficiency in high-quality evidence for or against the use of ARPs in the treatment of GERD may account for some of the variation in care seen in this population. Goldin et al. examined the use of ARPs at free-standing pediatric hospitals between 2001 and 2005 and found a decrease in overall ARP rates annually. However, when individual hospitals were examined, there was considerable interhospital variability with some hospitals performing significantly greater and some significantly fewer ARPs.24 The authors suggest that this finding could be due to the existing variability in the indications for ARPs. Without clear guidelines for when to perform ARPs, physicians are allowed more latitude when it comes to the decision of who gets ARPs and who does not, and decisions are made based on factors such as surgeon preference and anecdotal evidence rather than high-quality outcomes research. For instance, a surgeon who
experiences a mortality associated with an aspiration event in an infant might choose to perform an ARP on all patients who present with an aspiration pneumonia. A practice such as this could easily lead to overutilization of ARPs in children; however, it is impossible to determine whether surgery is being overutilized or underutilized if there are no well-defined indications for surgery. It is possible that individual pediatric surgery groups or even individual surgeons who are high utilizers of ARPs are responsible for the high state utilization rates seen in some states, especially those that have utilization rates that do not correlate with the regional rates such as Vermont where the state utilization rate is high within the overall low utilization rate of the Northeast. To reduce variability in the use of ARPs in children, there are several areas for potential improvement and opportunity for further research. First, the indications for surgery need to be clarified. A standardized diagnostic approach and trial period for medical management before surgical referral need
s h o r t e t a l v a r i a b i l i t y i n a n t i r e fl u x s u r g e r y u s e
to be established and adopted by gastroenterologists and surgeons. In addition, high-quality evidence for the use of ARPs over medical management in the form of a randomized trial needs to be ascertained especially in infants where GERD often resolves spontaneously overtime. Finally, the development of a decision analysis tool for the use of ARPs in children would allow physicians to use the best available evidence with probabilities of outcomes to select the best treatment option for patients.17 There are several limitations to this study. First, the use of an administrative database limits our analysis to only those variables included in the database. There may be key variables driving ARP utilization that we were unable to assess. We were unable to assess the diagnostic work-up that each patient received before surgical intervention and therefore could not determine what parameters were used as an indication for surgery. Furthermore, we were unable to follow patients after surgery to determine if the ARP conferred benefit. The KID only includes 44 of the 50 states in the United States, so we were unable to analyze data from all states. However, the methods used for generating national estimates should allow for an accurate representation of the entire United States. We only analyzed 1 y of KID data; so we were not able to establish any trends over time in the geographic variability of ARP utilization. Finally, the KID does not contain provider level data, and we were unable to analyze provider-specific variation.
Conclusions Significant regional and state-level variation in ARP utilization exists that cannot be explained entirely by differences in patient age, patient raceeethnicity, case-mix, and surgical approach. To decrease variation in care and potentially improve care, further research is warranted to delineate local factors driving surgical utilization of ARP in children. Consensus guidelines regarding indications and appropriateness for use are needed.
Acknowledgment This work was supported by the Emory þ Children’s Pediatric Research Center and Children’s Healthcare of Atlanta. Authors’ contributions: H.S. is the first author and contributed to conception and design, acquisition of data, interpretation of the data, drafting of the article, and final approval of the article. W.Z. participated in concept and design, acquisition of the data, analysis and interpretation of the data, drafting and critical revision of the article, and final approval of the article. C.M., C.T., L.A.W., and M.V.R. participated in concept and design, interpretation of the data, critical revision, and final approval of the article. All authors are responsible for the accuracy and integrity of the work.
Disclosures The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
55
references
1. Vandenplas Y, Rudolph CD, Di Lorenzo C, et al. Pediatric gastroesophageal reflux clinical practice guidelines: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN). J Pediatr Gastroenterol Nutr. 2009;49:498e547. 2. Hegar B, Dewanti NR, Kadim M, Alatas S, Firmansyah A, Vandenplas Y. Natural evolution of regurgitation in healthy infants. Acta Paediatr. 2009;98:1189e1193. 3. Hassall E. Decisions in diagnosing and managing chronic gastroesophageal reflux disease in children. J Pediatr. 2005;146:S3eS12. 4. Hassall E. Antireflux surgery in children: time for a harder look. Pediatrics. 1998;101:467e468. 5. Di Lorenzo C, Orenstein S. Fundoplication: friend or foe? J Pediatr Gastroenterol Nutr. 2002;34:117e124. 6. Nusrat S, Bielefeldt K. Fundoplication for gastroesophageal reflux disease: regional variability and factors predicting operative approach. Dis Esophagus. 2014;27:719e725. 7. Finlayson SR, Birkmeyer JD, Laycock WS. Trends in surgery for gastroesophageal reflux disease: the effect of laparoscopic surgery on utilization. Surgery. 2003;133:147e153. 8. Finlayson SR, Laycock WS, Birkmeyer JD. National trends in utilization and outcomes of antireflux surgery. Surg Endosc. 2003;17:864e867. 9. McMahon RL, Mercer CD. National trends in gastroesophageal reflux surgery. Can J Surg. 2000;43:48e52. 10. Finks JF, Wei Y, Birkmeyer JD. The rise and fall of antireflux surgery in the United States. Surg Endosc. 2006;20:1698e1701. 11. Funk LM, Kanji A, Scott Melvin W, Perry KA. Elective antireflux surgery in the US: an analysis of national trends in utilization and inpatient outcomes from 2005 to 2010. Surg Endosc. 2014;28:1712e1719. 12. Lasser MS, Liao JG, Burd RS. National trends in the use of antireflux procedures for children. Pediatrics. 2006;118:1828e1835. 13. Introduction to the HCUP KID’s Inpatient database (KID). Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality; 2012. Available at: https://www.hcup-us.ahrq.gov/db/nation/kid/KID_2012_ Introduction.pdf, Accessed March 21, 2017. 14. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2015. Available at: https://www.R-project.org/, Accessed March 21, 2017. 15. Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1e48. 16. Birkmeyer JD, Reames BN, McCulloch P, Carr AJ, Campbell WB, Wennberg JE. Understanding of regional variation in the use of surgery. Lancet. 2013;382:1121e1129. 17. McCulloch P, Nagendran M, Campbell WB, et al. Strategies to reduce variation in the use of surgery. Lancet. 2013;382:1130e1139. 18. International Pediatric Endosurgery G. IPEG guidelines for the surgical treatment of pediatric gastroesophageal reflux disease (GERD). J Laparoendosc Adv Surg Tech A. 2009;19(Suppl 1). x-xiii. 19. LaRiviere CA, Parimi C, Huaco JC, Acierno SA, Garrison MM, Goldin AB. Variations in preoperative decision making for antireflux procedures in pediatric gastroesophageal reflux disease: a survey of pediatric surgeons. J Pediatr Surg. 2011;46:1093e1098.
56
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
20. Gilger MA, Yeh C, Chiang J, Dietrich C, Brandt ML, El-Serag HB. Outcomes of surgical fundoplication in children. Clin Gastroenterol Hepatol. 2004;2:978e984. 21. Fonkalsrud EW, Ashcraft KW, Coran AG, et al. Surgical treatment of gastroesophageal reflux in children: a combined hospital study of 7467 patients. Pediatrics. 1998;101:419e422. 22. Mathei J, Coosemans W, Nafteux P, et al. Laparoscopic Nissen fundoplication in infants and children: analysis of 106
consecutive patients with special emphasis in neurologically impaired vs. neurologically normal patients. Surg Endosc. 2008;22:1054e1059. 23. Norrashidah AW, Henry RL. Fundoplication in children with gastro-oesophageal reflux disease. J Paediatr Child Health. 2002;38:156e159. 24. Goldin AB, Garrison M, Christakis D. Variations between hospitals in antireflux procedures in children. Arch Pediatr Adolesc Med. 2009;163:658e663.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Association for Academic Surgery
Exploring regional variability in utilization of antireflux surgery in children Heather L. Short, MD,a Wanzhe Zhu, BS,b Courtney McCracken, PhD,c Curtis Travers, MPH,c Lance A. Waller, PhD,b and Mehul V. Raval, MD, MSa,* a
Division of Pediatric Surgery, Department of Surgery, Children’s Healthcare of Atlanta, Emory University School of Medicine, Atlanta, Georgia b Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University School of Medicine, Atlanta, Georgia c Department of Pediatrics, Emory University School of Medicine, Atlanta, Georgia
article info
abstract
Article history:
Background: There is significant variation surrounding the indications, surgical approaches,
Received 2 December 2016
and outcomes for children undergoing antireflux procedures (ARPs) resulting in geographic
Received in revised form
variation of care. Our purpose was to quantify this geographic variation in the utilization of
12 January 2017
ARPs in children.
Accepted 28 February 2017
Methods: A cross-sectional analysis of the 2009 Kid’s Inpatient Database was performed to
Available online 8 March 2017
identify patients with gastroesophageal reflux disease or associated diagnoses. Regional surgical utilization rates were determined, and a mixed effects model was used to identify
Keywords:
factors associated with the use of ARPs.
Pediatric surgery
Results: Of the 148,959 patients with a diagnosis of interest, 4848 (3.3%) underwent an ARP
Antireflux surgery
with 2376 (49%) undergoing a laparoscopic procedure. The Northeast (2.0%) and Midwest
Variation in care
(2.2%) had the lowest overall utilization of surgery, compared with the South (3.3%) and
GERD
West (3.4%). After adjustment for age, case-mix, and surgical approach, variation persisted
Regional variation
with the West and the South demonstrating almost two times the odds of undergoing an ARP compared with the Northeast. Surgical utilization rates are independent of state-level volume with some of the highest case volume states having surgical utilization rates below the national rate. In the West, the use of laparoscopy correlated with overall utilization of surgery, whereas surgical approach was not correlated with ARP use in the South. Conclusions: Significant regional variation in ARP utilization exists that cannot be explained entirely by differences in patient age, race/ethnicity, case-mix, and surgical approach. In order to decrease variation in care, further research is warranted to establish consensus guidelines regarding indications for the use ARPs for children. ª 2017 Elsevier Inc. All rights reserved.
Institutional Review Board approval: Exempt. * Corresponding author. Division of Pediatric Surgery, Department of Surgery, Children’s Healthcare of Atlanta, Emory University School of Medicine, 1405 Clifton Road NE, Atlanta, GA 30322. Tel.: þ1 404-785-0781; fax: þ1 404 785-0800. E-mail address: [email protected] (M.V. Raval). 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.02.075
50
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
Introduction Gastroesophageal reflux is the normal physiological passage of gastric contents into the esophagus that can occur several times a day in infants and children.1 Most of these cases resolve spontaneously by 18 mo of life and do not cause any symptoms or complications.2 However, physiological reflux can progress to gastroesophageal reflux disease (GERD) and result in numerous pathologies that can negatively impact a child’s well-being. Medical management, specifically the use of proton pump inhibitors, is the mainstay of treatment for symptomatic reflux; however, antireflux procedures (ARPs) remain an option for children whose symptoms do not respond to medical management. Although there appear to be obvious benefits to the surgical treatment of GERD, there is significant variation in the
indications, preoperative evaluation, surgical approach, and outcomes for children undergoing ARPs.3-5 Nusrat and Bielefeldt demonstrated a fivefold variation in regional ARP utilization rates in the United States for the treatment of GERD in adults.6 In addition, there was a marked increase in the use of ARPs for treatment of GERD in adults with the widespread introduction of laparoscopy in the 1990s.7-9 National surgical utilization rates declined steadily from 1999 to 200310; however, the volume of elective adult ARPs has remained relatively stable since about 2005.11 Surprisingly, when a pediatric population was examined between 1996 and 2003, no similar trends were identified.12 The purpose of this study was to examine contemporary regional and state-level variability in utilization of ARPs to treat children with GERD and its complications. We hypothesized that regional and state level rates of surgical utilization would correlate with use of laparoscopy.
Table 1 e Demographics (age, race, and comorbidities), including all diagnoses with proportion that underwent ARPs. All patients
Total N Mean number of total diagnoses (SD)y
148,969
*
Antireflux surgery Yes
No
4230 (2.8%)
144,739 (97.2%)
7.672 (4.82)
9.752 (6.284)
7.611 (4.759)
71,089 (47.7%)
1939 (45.8%)
69,150 (47.8%)
69,938 (46.9%)
2009 (47.5%)
67,929 (46.9%)
Between 1 and 10 y
39,147 (26.3%)
1565 (37.0%)
37,582 (26.0%)
>10 y
39,468 (26.5%)
636 (15.0%)
38,832 (26.8%)
White
71,025 (47.7%)
2065 (48.8%)
68,960 (47.6%)
Black
21,201 (14.2%)
624 (14.8%)
20,577 (14.2%)
Female, N (%)z Age, N (%)x <1 y
Race, N (%)k
Hispanic
23,623 (15.9%)
668 (15.8%)
22,955 (15.9%)
Other
10,503 (7.1%)
332 (7.8%)
10,171 (7.0%)
26,028 (17.5%)
508 (12.0%)
25,520 (17.6%)
Region, N (%) Northeast Midwest
34,800 (23.4%)
763 (18.0%)
34,037 (23.5%)
South
56,457 (37.9%)
1884 (44.5%)
54,573 (37.7%)
West
31,684 (21.3%)
1075 (25.4%)
30,609 (21.1%)
4300 (2.9%)
184 (4.3%)
4116 (2.8%)
Contributing diagnoses (more than one may apply), N (%) Esophagitis, N (%) Esophageal stricture, N (%)
11,793 (7.9%)
Aspiration, N (%) Apnea, N (%) Failure to thrive, N (%) Barrett’s esophagus, N (%) Paraesophageal hernia/hiatal hernia, N (%) Esophageal reflux, N (%) Esophageal ulcer, N (%) *
828 (0.6%)
Dysphagia, N (%)
54 (1.3%)
774 (0.5%)
553 (13.1%)
11,240 (7.8%)
10,994 (7.4%)
414 (9.8%)
10,580 (7.3%)
8987 (6.0%)
143 (3.4%)
8844 (6.1%)
27,156 (18.2%)
1366 (32.3%)
25,790 (17.8%)
123 (0.1%)
23 (0.5%)
100 (0.1%)
2820 (1.9%)
597 (14.1%)
2223 (1.5%)
105,265 (70.7%) 445 (0.3%)
3840 (90.8%)
101,425 (70.1%)
13 (0.3%)
Of the 148,969 patients, 22,904 patients (15.4%) are missing either age or race. Number of diagnoses range from 1 to 35 and may contain other diagnoses not associated with ARP. z Gender is missing for 278 patients (0.2%). x Age is missing for 416 patients (0.3%). k Race is missing for 22,617 patients (15.2%), this includes all patients from Ohio, North Carolina, West Virginia, and Minnesota. y
432 (0.3%)
51
s h o r t e t a l v a r i a b i l i t y i n a n t i r e fl u x s u r g e r y u s e
Methods A cross-sectional analysis of the Agency for Healthcare Research and Quality (AHRQ) sponsored Healthcare Cost and Utilization Project (HCUP) Kids’ Inpatient Database (KID) was performed to identify children with a reflux associated diagnosis in 2009. The KID is a national administrative dataset containing hospital discharge-level data for patients aged <21 y. The dataset contains data from 44 states and more than 4100 US hospitals including information for over 3 million pediatric discharges per year sampled from all community, nonrehabilitation hospitals of states participating in HCUP. National estimates of discharges were generated using discharge weights provided by HCUP that are based on the poststratification of hospitals according to characteristics including ownership, bed size, teaching status, rural/urban location, region of the United States, and being a freestanding children’s hospital. Records from each hospital are randomly sampled to include 10% of uncomplicated births and 80% of complicated births because less variation in outcomes among uncomplicated births is expected. More than 100 clinical and nonclinical data elements are available for each discharge record including patient demographics, hospital charges, and diagnosis and procedure codes.13 Using the International Classification of Diseases, Ninth Revision, Clinical Manifestation, all patients (aged <21 years) with a diagnosis of GERD (530.81) or one of its related diagnoses, including esophagitis (530.10, 530.11), esophageal stricture (530.3), dysphagia (787.20, 787.21, 787.22, 787.23, 787.24), aspiration (507.0), apnea (786.03), failure to thrive (783.41, 779.34), Barrett’s esophagus (530.85), esophageal ulcer (530.20), or hiatal hernia (553.3) were identified. Children who underwent an ARP were identified with International Classification of Diseases, Ninth Revision, procedure codes for “other procedures for creation of esophagogastric sphincteric competence” (44.66) or “laparoscopic procedures for creation of esophagogastric sphincteric competence” (44.67). Demographics obtained from the database included age, race, number of diagnoses, location of hospital visit (hospital ID and state). Hospital-related information obtained from the database included hospital ID, state, and census region (Northeast, South, Midwest, and West). Individual state and regional surgical utilization rates were calculated by tabulating the number of ARPs performed for a given state or region, then dividing by the total number of records with at least one of the diagnoses identified above for that state. Descriptive statistics were calculated for all variables of interest and include means and standard deviations or counts and percentages, as appropriate. A mixed effects model was used to determine which regions and individual states were high utilizers of surgery and to identify patient and hospital factors associated with open versus laparoscopic procedures. A separate mixed effects logistic model was used to identify patient and hospital factors associated with the decision to perform ARPs on patients diagnosed with GERD or one of its related diagnoses, with the hospital’s state being the random effect. The addition of a random intercept improved model fit considerably over a fixed effect model as measured by
Table 2 e Factors associated with utilization of antireflux surgery based on a multivariate mixed-effects logistic model. OR (95% CI Lower limit-Upper limit)
P
Age <1 y
1.00
Reference
Between 1 and 10 y
1.23 (1.15-1.31)
<0.001
>10 y
0.48 (0.44-0.53)
<0.001
White
1.00
Reference
Black
1.01 (0.93-1.09)
0.900
Hispanic
0.87 (0.80-0.94)
<0.001
Race
Other Number of diagnoses*
1.02 (0.92-1.14)
0.686
1.07 (1.06-1.07)
<0.001
1.00
Reference
Region Northeast Midwest
1.30 (0.81-2.08)
0.271
South
1.97 (1.27-3.03)
0.002
West
1.85 (1.18-2.89)
0.007
Esophagitis
2.83 (2.39-3.34)
<0.001
Esophageal stricture
2.40 (1.75-3.30)
<0.001
Dysphagia
2.51 (2.29-2.76)
<0.001
Aspiration
2.65 (2.39-2.95)
<0.001
Apnea
1.20 (1.03-1.40)
0.022
Failure to thrive
4.27 (4.00-4.56)
<0.001
Barrett’s esophagus
9.20 (5.43-15.60)
<0.001
Paraesophageal/ hiatal hernia
22.23 (19.88-24.87)
<0.001
Esophageal reflux
10.93 (9.90-12.06)
<0.001
Esophageal ulcer
0.47 (0.24-0.91)
Contributing diagnoses
*
0.026
Number of total diagnoses ranges from 1 to 35.
Akaike’s Information Criterion. Results from the mixed effect models are presented as odds ratios with associated 95% confidence intervals (CIs). Analyses were performed in R 3.2.2 using the lme4 package function for mixed effects models.14,15 A P < 0.05 was considered statistically significant.
Results Of the 148,969 patients with one or more of the diagnoses of interest, 4230 (2.84%; 95% CI 2.76-2.92) patients underwent an ARP. Of the 4230 patients who underwent an ARP, 2095 (49.53%) underwent a laparoscopic procedure, 2125 (50.24%) underwent an open procedure, and 10 (0.23%) were indeterminate. Demographics of the cohort are presented in Table 1. Hispanics were less likely to undergo ARP compared with non-Hispanic whites (OR 0.87; 95% CI 0.80-0.94). GERD was the most common indication (91%) for ARPs (Table 1).
52
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
Fig. 1 e Panel 1. State-specific ratios of laparoscopic utilization rates over ARP utilization rates from the 2009 KID. Panel 2. Choropleth map of the United States demonstrating state-by-state laparoscopy utilization rates. States shaded in gray are not reported in the 2009 KID. (Color version of the figure is available online.)
Using the population aged <1 y as a reference, children aged between 1 and 10 y had increased odds of undergoing an ARP (OR 1.23; 95% CI 1.15-1.31). Older children (aged 11-20 y), however, had lower ARP utilization compared with children aged <1 y (OR 0.48; 95% CI 0.44-0.53). The number of diagnoses per patient also had a positive impact on ARP utilization. Each additional diagnosis increases the odds of a patient undergoing ARP by 7% (OR 1.07; 95% CI 1.06-1.07; Table 2). The Northeast (2.0%) and Midwest (2.2%) had the lowest overall utilization of surgery, compared with the South (3.3%) and West (3.4%). After adjustment for age, case-mix, and surgical approach and allowing for state specific utilization rates, regional variation persisted with the West and the South demonstrating close to two times the odds of undergoing an ARP compared with the Northeast (Table 2). Surgical utilization rates appeared to be independent of state-level case volume with some of the highest case volume states (California, New York, and Ohio) having surgical utilization
rates below the national rate (Fig. 1, Panel 1). There appears to be an overall bias toward ARP utilization in the West and South, whereas there is more state-to-state variability in the Northeast and Midwest (Fig. 1, Panel 2). Each of the diagnoses of interest was statistically significantly predictors of a patient undergoing an ARP. A patient diagnosed with esophagitis, esophageal stricture, dysphagia, or aspiration had about 2.5 times the odds of undergoing an ARP compared with patients who did not have that diagnosis. A patient diagnosed with Barrett’s esophagus or esophageal reflux had close to 10 times the odds of undergoing an ARP. A patient diagnosed with hernia or hiatal hernia had the highest odds of undergoing an ARP. In contrast, a patient diagnosed with apnea or esophageal ulcer had the lowest odds of undergoing an ARP (Table 2). The effect of these diagnoses have on ARP utilization is multiplicative so a patient diagnosed with aspiration and failure to thrive had over 11 times the odds of undergoing ARP than the same patient without these two diagnoses.
s h o r t e t a l v a r i a b i l i t y i n a n t i r e fl u x s u r g e r y u s e
Table 3 e Factors associated with the decision to perform laparoscopic procedure versus open procedure. Factor
Adjusted OR (95% CI Lower limit-Upper limit)
P
1.00
Reference
Age <1 y Between 1 and 10 y
1.21 (1.06-1.38)
0.005
>10 y
1.91 (1.59-2.31)
<0.001
White
1.00
Reference
Black
0.83 (0.70-0.98)
0.025
Hispanic
0.95 (0.80-1.12)
0.514
Other
0.89 (0.72-1.10)
0.290
0.94 (0.93-0.95)
<0.001
1.00
Reference
Race
Number of total diagnoses* Region Northeast Midwest
1.71 (0.80-3.63)
0.167
South
1.15 (0.57-2.31)
0.698
West
3.12 (1.51-6.48)
0.002
Dysphagia
1.23 (1.03-1.46)
0.019
Apnea
1.66 (1.21-2.28)
0.002
Failure to thrive
1.33 (1.17-1.50)
<0.001
*
Number of total diagnoses ranges from 1 to 35.
Many of the same factors associated with ARP utilization are also associated with the use of laparoscopic versus open technique. When undergoing ARP, children aged between 1 and 10 y have slightly higher odds of undergoing a laparoscopic procedure than a <1-year-old (OR 1.21; 95% CI 1.06-1.38). Children aged 10-20 y have twice the odds of undergoing a laparoscopic procedure (OR 1.91; 95% CI 1.59-2.31; Table 3). Each additional diagnosis also reduces the odds of a patient undergoing laparoscopy (OR 0.94; 95% CI 0.93-0.95). Using the Northeast as a reference, patients receiving care in the West have the highest odds of undergoing a laparoscopic procedure (OR 3.12; 95% CI 1.51-6.48; Table 3). Of the 10 diagnoses of interest, only dysphagia, apnea, and failure to thrive are associated with laparoscopy usage (Table 3). There is a weak but nonsignificant correlation (0.29; 95% CI 0.05-0.57; P ¼ 0.089) between laparoscopy use and state-level case volume, and we found considerable state-to-state variation in laparoscopy use within each region (Fig. 2).
Discussion Although the mainstay of treatment for GERD and its associated diagnoses is medical management, ARPs remain an option for medically refractory cases in children. There is significant variation surrounding the indications, surgical approaches, and outcomes in children undergoing ARPs resulting in geographical variation in the utilization of surgery in the United States. Significant regional variation in ARP utilization exists with the West and the South demonstrating
53
the highest surgical utilization rates, and adjustment for age, case-mix, and surgical approach did not explain this variation. In addition, surgical utilization rates were found to be independent of state-level case volumes with some of the states with the highest case volume demonstrating surgical utilization rates below the national average. Although we hypothesized that the availability of laparoscopy and surgeons proficient in laparoscopic technique could be a driver for ARP utilization, we did not find a significant correlation between surgical utilization rates and the use laparoscopy. Rather, it appears that ARP utilization rates vary more on the state level as opposed to the regional level, and case volume does not appear to drive laparoscopy utilization, which varies considerably from state-to-state, as well. Regional variation in the use of surgical procedures is often attributed to differences in physician beliefs about the indications for surgery and patient preferences with differences in illness prevalence playing only a small role. This explanation can also be applied to the state-by-state variation that we saw in our results and may even be applicable to variation among individual hospitals. Practice variation exists in other areas of medicine, but the implication of variation in when to perform an operation or not to, is that some patients may be getting procedures that they do not want or need.16 Ideally, surgical treatment should provide maximal benefit to the patient with minimal risk of complications. When surgery is overutilized, or used in a patient who is already receiving maximal benefit from nonsurgical management, that patient is put at risk unnecessarily and the system is incurring unnecessary costs. Variation in care is a driver for systemic inefficiency, and reducing variation may decrease complexity of care, encourage adherence to best practices, generate cost savings, and lead to improved outcomes for some procedures.17 Recommendations from the International Pediatric Endosurgery Group state that surgery is indicated as a treatment for GERD when the diagnosis has been confirmed by objective testing such as pH monitoring, impedance monitoring, or esophagoscopy.18 Despite this, a survey of pediatric surgeons working in hospitals that are members of the Children’s Health Care Association found that only 13% of those surveyed require a pH probe study preoperatively.19 Furthermore, it is recommended that only children who fail medical treatment of their GERD be considered for an ARP; however, many children are not trialed on medical therapy before surgery, and there are no randomized trials comparing medical therapy alone with fundoplication.1 The definition of failure of medical management remains vague and often the decision to perform ARPs is based on the presence of clinical symptoms such as pain, aspiration, failure to thrive, apnea, or acute lifethreatening events in the neonate. These factors allow for wide variation in the clinical decision making in regards to the utilization of surgery in the treatment of GERD in children. Most of the existing literature on outcomes in the surgical treatment of GERD consists of retrospective case series in which the documentation of diagnostic testing and the details of previous medical management are lacking.20-22 Moreover, many of the studies that describe positive outcomes after ARPs use subjective measures of outcome to assess their efficacy rather than objective tests to confirm the resolution of
54
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
Fig. 2 e Panel 1. State-specific ratios of ARP utilization rates from the 2009 KID demonstrating that surgical utilization rates do not correlate with the volume of patients with a diagnosis of interest. Panel 2. Choropleth map of the United States demonstrating state-by-state ARP utilization rates per 1000 diagnosis of interest. States shaded in gray are not reported in the 2009 KID. (Color version of the figure is available online.)
symptoms.21,23 This makes it difficult to come to a meaningful conclusion regarding postoperative outcomes in children undergoing ARPs. The deficiency in high-quality evidence for or against the use of ARPs in the treatment of GERD may account for some of the variation in care seen in this population. Goldin et al. examined the use of ARPs at free-standing pediatric hospitals between 2001 and 2005 and found a decrease in overall ARP rates annually. However, when individual hospitals were examined, there was considerable interhospital variability with some hospitals performing significantly greater and some significantly fewer ARPs.24 The authors suggest that this finding could be due to the existing variability in the indications for ARPs. Without clear guidelines for when to perform ARPs, physicians are allowed more latitude when it comes to the decision of who gets ARPs and who does not, and decisions are made based on factors such as surgeon preference and anecdotal evidence rather than high-quality outcomes research. For instance, a surgeon who
experiences a mortality associated with an aspiration event in an infant might choose to perform an ARP on all patients who present with an aspiration pneumonia. A practice such as this could easily lead to overutilization of ARPs in children; however, it is impossible to determine whether surgery is being overutilized or underutilized if there are no well-defined indications for surgery. It is possible that individual pediatric surgery groups or even individual surgeons who are high utilizers of ARPs are responsible for the high state utilization rates seen in some states, especially those that have utilization rates that do not correlate with the regional rates such as Vermont where the state utilization rate is high within the overall low utilization rate of the Northeast. To reduce variability in the use of ARPs in children, there are several areas for potential improvement and opportunity for further research. First, the indications for surgery need to be clarified. A standardized diagnostic approach and trial period for medical management before surgical referral need
s h o r t e t a l v a r i a b i l i t y i n a n t i r e fl u x s u r g e r y u s e
to be established and adopted by gastroenterologists and surgeons. In addition, high-quality evidence for the use of ARPs over medical management in the form of a randomized trial needs to be ascertained especially in infants where GERD often resolves spontaneously overtime. Finally, the development of a decision analysis tool for the use of ARPs in children would allow physicians to use the best available evidence with probabilities of outcomes to select the best treatment option for patients.17 There are several limitations to this study. First, the use of an administrative database limits our analysis to only those variables included in the database. There may be key variables driving ARP utilization that we were unable to assess. We were unable to assess the diagnostic work-up that each patient received before surgical intervention and therefore could not determine what parameters were used as an indication for surgery. Furthermore, we were unable to follow patients after surgery to determine if the ARP conferred benefit. The KID only includes 44 of the 50 states in the United States, so we were unable to analyze data from all states. However, the methods used for generating national estimates should allow for an accurate representation of the entire United States. We only analyzed 1 y of KID data; so we were not able to establish any trends over time in the geographic variability of ARP utilization. Finally, the KID does not contain provider level data, and we were unable to analyze provider-specific variation.
Conclusions Significant regional and state-level variation in ARP utilization exists that cannot be explained entirely by differences in patient age, patient raceeethnicity, case-mix, and surgical approach. To decrease variation in care and potentially improve care, further research is warranted to delineate local factors driving surgical utilization of ARP in children. Consensus guidelines regarding indications and appropriateness for use are needed.
Acknowledgment This work was supported by the Emory þ Children’s Pediatric Research Center and Children’s Healthcare of Atlanta. Authors’ contributions: H.S. is the first author and contributed to conception and design, acquisition of data, interpretation of the data, drafting of the article, and final approval of the article. W.Z. participated in concept and design, acquisition of the data, analysis and interpretation of the data, drafting and critical revision of the article, and final approval of the article. C.M., C.T., L.A.W., and M.V.R. participated in concept and design, interpretation of the data, critical revision, and final approval of the article. All authors are responsible for the accuracy and integrity of the work.
Disclosures The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
55
references
1. Vandenplas Y, Rudolph CD, Di Lorenzo C, et al. Pediatric gastroesophageal reflux clinical practice guidelines: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN). J Pediatr Gastroenterol Nutr. 2009;49:498e547. 2. Hegar B, Dewanti NR, Kadim M, Alatas S, Firmansyah A, Vandenplas Y. Natural evolution of regurgitation in healthy infants. Acta Paediatr. 2009;98:1189e1193. 3. Hassall E. Decisions in diagnosing and managing chronic gastroesophageal reflux disease in children. J Pediatr. 2005;146:S3eS12. 4. Hassall E. Antireflux surgery in children: time for a harder look. Pediatrics. 1998;101:467e468. 5. Di Lorenzo C, Orenstein S. Fundoplication: friend or foe? J Pediatr Gastroenterol Nutr. 2002;34:117e124. 6. Nusrat S, Bielefeldt K. Fundoplication for gastroesophageal reflux disease: regional variability and factors predicting operative approach. Dis Esophagus. 2014;27:719e725. 7. Finlayson SR, Birkmeyer JD, Laycock WS. Trends in surgery for gastroesophageal reflux disease: the effect of laparoscopic surgery on utilization. Surgery. 2003;133:147e153. 8. Finlayson SR, Laycock WS, Birkmeyer JD. National trends in utilization and outcomes of antireflux surgery. Surg Endosc. 2003;17:864e867. 9. McMahon RL, Mercer CD. National trends in gastroesophageal reflux surgery. Can J Surg. 2000;43:48e52. 10. Finks JF, Wei Y, Birkmeyer JD. The rise and fall of antireflux surgery in the United States. Surg Endosc. 2006;20:1698e1701. 11. Funk LM, Kanji A, Scott Melvin W, Perry KA. Elective antireflux surgery in the US: an analysis of national trends in utilization and inpatient outcomes from 2005 to 2010. Surg Endosc. 2014;28:1712e1719. 12. Lasser MS, Liao JG, Burd RS. National trends in the use of antireflux procedures for children. Pediatrics. 2006;118:1828e1835. 13. Introduction to the HCUP KID’s Inpatient database (KID). Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality; 2012. Available at: https://www.hcup-us.ahrq.gov/db/nation/kid/KID_2012_ Introduction.pdf, Accessed March 21, 2017. 14. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2015. Available at: https://www.R-project.org/, Accessed March 21, 2017. 15. Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1e48. 16. Birkmeyer JD, Reames BN, McCulloch P, Carr AJ, Campbell WB, Wennberg JE. Understanding of regional variation in the use of surgery. Lancet. 2013;382:1121e1129. 17. McCulloch P, Nagendran M, Campbell WB, et al. Strategies to reduce variation in the use of surgery. Lancet. 2013;382:1130e1139. 18. International Pediatric Endosurgery G. IPEG guidelines for the surgical treatment of pediatric gastroesophageal reflux disease (GERD). J Laparoendosc Adv Surg Tech A. 2009;19(Suppl 1). x-xiii. 19. LaRiviere CA, Parimi C, Huaco JC, Acierno SA, Garrison MM, Goldin AB. Variations in preoperative decision making for antireflux procedures in pediatric gastroesophageal reflux disease: a survey of pediatric surgeons. J Pediatr Surg. 2011;46:1093e1098.
56
j o u r n a l o f s u r g i c a l r e s e a r c h 1 5 j u n e 2 0 1 7 ( 2 1 4 ) 4 9 e5 6
20. Gilger MA, Yeh C, Chiang J, Dietrich C, Brandt ML, El-Serag HB. Outcomes of surgical fundoplication in children. Clin Gastroenterol Hepatol. 2004;2:978e984. 21. Fonkalsrud EW, Ashcraft KW, Coran AG, et al. Surgical treatment of gastroesophageal reflux in children: a combined hospital study of 7467 patients. Pediatrics. 1998;101:419e422. 22. Mathei J, Coosemans W, Nafteux P, et al. Laparoscopic Nissen fundoplication in infants and children: analysis of 106
consecutive patients with special emphasis in neurologically impaired vs. neurologically normal patients. Surg Endosc. 2008;22:1054e1059. 23. Norrashidah AW, Henry RL. Fundoplication in children with gastro-oesophageal reflux disease. J Paediatr Child Health. 2002;38:156e159. 24. Goldin AB, Garrison M, Christakis D. Variations between hospitals in antireflux procedures in children. Arch Pediatr Adolesc Med. 2009;163:658e663.