Complications and Antireflux Medication Use After Antireflux Surgery
2006; Elsevier BV; Volume: 4; Issue: 3 Linguagem: Inglês
10.1016/j.cgh.2005.12.019
ISSN1542-7714
AutoresJason A. Dominitz, Christopher A. Dire, Kevin G. Billingsley, Jeffrey A. Todd–Stenberg,
Tópico(s)Helicobacter pylori-related gastroenterology studies
ResumoBackground & Aims: Although antireflux surgery is increasingly common, few studies have assessed the associated complications and health care use after surgery. The aim of this study was to estimate postoperative complications and continued use of antireflux medications and to identify predictors of complications. Methods: Through a review of the Department of Veterans Affairs administrative databases, all patients undergoing antireflux surgery from October 1, 1990, through January 29, 2001, were identified. Of 3367 patients identified, 222 were excluded as a result of a diagnosis of esophageal cancer, achalasia, or because there was no diagnosis related to gastroesophageal reflux disease. Medication use was determined for 2406 patients who had a minimum of 1 year of follow-up, including 1 or more outpatient visits at least 6 months after surgery and during the time when national pharmacy records were available. Results: Dysphagia was recorded in 19.4%, dilation was performed in 6.4%, and a repeat antireflux surgery was performed in 2.3%. The surgical mortality rate was .8%. Prescriptions were dispensed repeatedly for H2 receptor antagonists in 23.8%, proton pump inhibitors in 34.3%, and promotility agents in 9.2% of patients. Overall, 49.8% of patients received at least 3 prescriptions for one of these medications. Conclusions: A moderate proportion of patients undergoing antireflux surgeries experienced complications and approximately 50% of patients received multiple prescriptions for antireflux medications at a median of 5 years of follow-up evaluation. Therefore, before surgery is performed, patients considering surgery should be counseled fully about the risk for complications and the likelihood of continued antireflux medication use. Background & Aims: Although antireflux surgery is increasingly common, few studies have assessed the associated complications and health care use after surgery. The aim of this study was to estimate postoperative complications and continued use of antireflux medications and to identify predictors of complications. Methods: Through a review of the Department of Veterans Affairs administrative databases, all patients undergoing antireflux surgery from October 1, 1990, through January 29, 2001, were identified. Of 3367 patients identified, 222 were excluded as a result of a diagnosis of esophageal cancer, achalasia, or because there was no diagnosis related to gastroesophageal reflux disease. Medication use was determined for 2406 patients who had a minimum of 1 year of follow-up, including 1 or more outpatient visits at least 6 months after surgery and during the time when national pharmacy records were available. Results: Dysphagia was recorded in 19.4%, dilation was performed in 6.4%, and a repeat antireflux surgery was performed in 2.3%. The surgical mortality rate was .8%. Prescriptions were dispensed repeatedly for H2 receptor antagonists in 23.8%, proton pump inhibitors in 34.3%, and promotility agents in 9.2% of patients. Overall, 49.8% of patients received at least 3 prescriptions for one of these medications. Conclusions: A moderate proportion of patients undergoing antireflux surgeries experienced complications and approximately 50% of patients received multiple prescriptions for antireflux medications at a median of 5 years of follow-up evaluation. Therefore, before surgery is performed, patients considering surgery should be counseled fully about the risk for complications and the likelihood of continued antireflux medication use. Gastroesophageal reflux disease (GERD) is a common condition affecting up to 20% of Americans on a weekly basis.1Locke 3rd, G.R. Talley N.J. Fett S.L. et al.Prevalence and clinical spectrum of gastroesophageal reflux a population-based study in Olmsted County, Minnesota.Gastroenterology. 1997; 112: 1448-1456Abstract Full Text PDF PubMed Scopus (1932) Google Scholar, 2The Gallup OrganizationA Gallup Organization national survey. The Gallup Organization, Princeton1988Google Scholar The total direct costs of GERD including all physician services, facility costs, and drug costs were estimated to be $9.3 billion in 1998.3Sandler R.S. Everhart J.E. Donowitz M. et al.The burden of selected digestive diseases in the United States.Gastroenterology. 2002; 122: 1500-1511Abstract Full Text Full Text PDF PubMed Scopus (1238) Google Scholar Although GERD usually can be managed with medical therapy (eg, acid suppression), some patients undergo surgical treatment for a variety of reasons, such as the failure of medications to control symptoms or the desire to avoid long-term medication use. The number of adult antireflux procedures performed in the United States more than tripled from 11,000 per year in 19854Pokras R. National Health Survey: detailed diagnosis and procedures for patients discharged from short-stay hospitals, United States, 1985. National Center for Health Statistics.Vital Health Stat 13. 1987; 90: 1-290PubMed Google Scholar to 40,000 in 2001,5Kozak L.J. Owings M.F. Hall M.J. National Hospital Discharge Survey 2001 annual summary with detailed diagnosis and procedure data. National Center for Health Statistics.Vital Health Stat 13. 2004; 156: 1-198PubMed Google Scholar at a population-based annual rate of 12.0 per 100,000 adults in 1997.6Finlayson S.R. Laycock W.S. Birkmeyer J.D. National trends in utilization and outcomes of antireflux surgery.Surg Endosc. 2003; 17: 864-867Crossref PubMed Scopus (117) Google Scholar This surgery has been associated with improvements in patients’ health-related quality of life.7Duffy J.P. Maggard M. Hiyama D.T. et al.Laparoscopic Nissen fundoplication improves quality of life in patients with atypical symptoms of gastroesophageal reflux.Am Surg. 2003; 69: 833-838PubMed Google Scholar, 8Trus T.L. Laycock W.S. Waring J.P. et al.Improvement in quality of life measures after laparoscopic antireflux surgery.Ann Surg. 1999; 229: 331-336Crossref PubMed Scopus (114) Google Scholar Although medical therapies typically are tested through large-scale, randomized, controlled trials, surgical therapies rarely undergo this evaluation. However, in 1992 the Department of Veterans Affairs (VA) conducted a randomized study comparing medical therapy with fundoplication surgery for 247 patients with complicated GERD.9Spechler S.J. The Department of Veterans Affairs Gastroesophageal Reflux Disease Study GroupComparison of medical and surgical therapy for complicated gastroesophageal reflux disease in veterans.N Engl J Med. 1992; 326: 786-792Crossref PubMed Scopus (519) Google Scholar Initial results indicated that surgery was found to be significantly better than medical therapy for controlling the symptoms of GERD, although a follow-up study determined that 62% of these patients were using antireflux medications regularly 6 years after their surgery.10Spechler S.J. Lee E. Ahnen D. et al.Long-term outcome of medical and surgical therapies for gastroesophageal reflux disease follow-up of a randomized controlled trial.JAMA. 2001; 285: 2331-2338Crossref PubMed Scopus (851) Google Scholar This study, similar to most others, originated from centers with specialized expertise and interest in this area. There are few studies published to date that evaluate fundoplication in the community setting. The purpose of this study was to estimate health care use and complications after antireflux surgery in all VA medical centers nationwide because the surgical experience may more closely reflect community hospitals than highly specialized tertiary care facilities. We obtained all data from the VA national Patient Treatment File (PTF), outpatient clinic file, National VA Pharmacy Database, and the Beneficiary Identification and Record Locator System for October 1, 1990, through February 28, 2001. The PTF contains diagnosis codes for all admissions at any VA medical center nationwide. For each discharge, up to 10 diagnoses, 5 operating room procedures, and 32 non–operating room procedures are listed according to the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM). Each record also contains demographic and administrative characteristics such as race, sex, age, eligibility for care at the VA center, marital status, and an identifier for the discharging hospital. The outpatient clinic file contains information on the date of the clinic visit, the clinic visited, diagnosis (ICD-9-CM codes since fiscal year 1997), and procedures (listed as current procedural terminology11Physicians’ current procedural terminology. American Medical Association, Chicago1993Google Scholar codes) performed during the visit. Prescriptions for medications including proton pump inhibitors (PPI), H2 receptor antagonists (H2RA), and promotility agents were identified through the recently created National VA Pharmacy Database. This database contains all prescriptions dispensed at any VA pharmacy nationwide since October 1, 1998. The Beneficiary Identification and Record Locator System database was examined to determine each veteran’s mortality status. The Beneficiary Identification and Record Locator System is a VA database created as an administrative tool to track the payment of benefits to veterans and their beneficiaries. This file contains the veteran’s name, Social Security number, and date of death (if applicable). The Beneficiary Identification and Record Locator System has been found to compare favorably with the National Death Index, with an estimated mortality ascertainment rate (94.5%), similar to that of the National Death Index (96.7%),12Fisher S. Weber L. Goldberg J. et al.Mortality ascertainment in the veteran population alternatives to the National Death Index.Am J Epidemiol. 1995; 141: 242-250PubMed Google Scholar and is particularly accurate for veterans who were inpatients.13Dominitz J.A. Maynard C. Boyko E.J. Assessment of vital status in Department of Veterans Affairs national databases. comparison with state death certificates.Ann Epidemiol. 2001; 11: 286-291Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar By using the PTF, we identified 3367 patients who have undergone an antireflux procedure from October 1, 1990, to January 29, 2001, using ICD-9-CM code 44.66 (other procedures for creation of esophagogastric sphincteric competence). To exclude prevalent cancers, patients with a diagnosis of esophageal cancer (ICD-9 150.0–151.0) within 1 year of the initial surgery were excluded (n = 61). Patients with achalasia (ICD-9 530.0) at any time in the study interval also were excluded (n = 82). To determine more accurately that these patients underwent antireflux surgery, we limited our analysis to those patients with a GERD-related diagnosis at the time of the index surgery (ICD-9 codes: 530.11, reflux esophagitis; 530.1x, esophagitis: not otherwise specified, chemical, peptic postoperative regurgitant, other, abscess of esophagus, 530.2x, ulcer of esophagus; 530.3x, stricture and stenosis of esophagus; 530.81, esophageal reflux [nonerosive esophagitis]). As a result, an additional 79 patients were excluded, leaving 3145 patients available for analysis of complications with a minimum of 30 days of follow-up evaluation. Because of the later creation of the pharmacy database, 2 additional study cohorts were evaluated. All patients without any of the earlier-described exclusions were included in the study of complications. Inclusion in the medication cohort required the following: (1) a minimum of 1 outpatient visit at least 6 months after surgery (to ensure continued VA care), (2) at least 1 outpatient clinic visit from October 1, 1998, through February 28, 2001 (the only period when pharmacy records were available), and (3) a minimum of 1 year of follow-up after their surgery (to allow for determination of long-term medication use). Therefore, patients undergoing surgery after February 28, 2000 (n = 259), were excluded from the medication cohort. Of the remaining 2886 patients, 212 died before October 1, 1998, and an additional 268 were not seen at any VA facility during the pharmacy period, leaving 2406 in the analytic cohort. A recent cohort also was established that included only those patients who had surgery between October 1, 1998, and February 28, 2000 (n = 404), allowing for complete pharmacy records of medication use in the perioperative period. Complications related to antireflux surgery were defined as dysphagia (ICD-9 787.2), esophageal dilation (ICD-9 42.92, current procedural terminology 43220, 43226, 43453, 43450), redo-fundoplication (ICD-9 44.66, excluding the first 30 days), and 30-day postoperative mortality. Because the outpatient clinic file database did not include diagnoses until fiscal year 1997, the occurrence of dysphagia could be determined only from inpatient records before that time. However, esophageal dilations that occurred either as an inpatient or an outpatient could be ascertained throughout the entire study period. We assessed the association between complications and comorbidity using the Charlson comorbidity index adapted for administrative databases.14Deyo R. Cherkin D. Ciol M. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases.J Clin Epidemiol. 1992; 45: 613-619Abstract Full Text PDF PubMed Scopus (8849) Google Scholar This index assigns a score for each patient determined by diagnostic codes found in administrative databases such as the PTF. We used ICD-9-CM codes from the 5 years before and including the index admission to determine the Charlson score. The measurement of antireflux medications included prescriptions for antacids (eg, calcium carbonate), H2RAs (ie, famotidine, ranitidine, nizatidine, cimetidine), PPIs (ie, lansoprazole, omeprazole, rabeprazole, pantoprazole), or promotility agents (ie, propulsid, metoclopramide). The primary medication outcome of interest was the chronic use of these medications. Chronic use was defined as 3 or more prescriptions for any medication within the category, including at least 1 prescription occurring no less than 6 months after the date of surgery. Because most prescriptions cover a 30- to 90-day supply, this definition reflects at least 3 months of medications for most patients. The proportion of patients with complications and using antireflux medications was calculated. Cox proportional hazards models were used to identify predictors of dysphagia, dilation, and redo-fundoplication after testing the proportional hazards assumption. A multivariate logistic regression model was used to identify predictors of surgical mortality. Covariates in the model included age at the time of surgery, race (coded as white or other), sex, hospital volume of antireflux surgeries during the study period (using the log of the number of surgeries), and era of the surgery (dichotomized as pre-1997 or later). The likelihood ratio test was used to evaluate the significance of each variable in the models.15Kleinbaum D. Kupper L. Muller K. Applied regression analysis and other multivariable methods. Duxbury Press, Belmont, CA1988Google Scholar In a secondary analysis we compared the surgical eras (ie, pre-1997 vs later) with respect to patient demographics and surgical mortality using χ2 and t tests. All analyses were performed using SAS software (version 9.1; SAS, Cary, NC). All P values were 2-sided, using α = .05 as the reference standard for determining significance. The Human Subjects Division of the University of Washington and the VA Puget Sound Health Care System Research Committee approved the study. Among the 3145 patients in the overall cohort, the median follow-up period was 4.5 years (range, 0–10.4 y). Table 1 shows patient demographics. Overall, 368 (11.7%) patients died during the study period. The proportions of patients reported to have dysphagia, dilations, redo-fundoplication, and surgical mortality and the timing of these complications are shown in Table 2. By using Cox models, increasing age was associated with an increased risk for having a diagnosis of dysphagia, undergoing esophageal dilation, and dying within 30 days of the surgery (Table 3). There was no association between race and these outcomes. However, men were less likely to have a diagnosis of dysphagia after surgery. Although recent surgeries had a significantly increased hazard ratio for a diagnosis of dysphagia, this finding likely was confounded by the lack of outpatient diagnoses before fiscal year 1997. There was no significant association between the volume of antireflux surgeries performed at a VA facility and each of these complications. When comparing surgeries performed before or after the beginning of 1997, recent surgical patients (n = 1272) were younger (52.3 vs 53.7 y, P = .009), less likely to be men (92% vs 96%, P < .0001), and less likely to be white (85% vs 90%, P < .0001) than the earlier surgical group. The surgical mortality rate was similar for recent compared with earlier surgical groups (.75% vs .94%, P = not significant). Although comorbidity scores were low overall, those who died had significantly higher comorbidity than those who survived (median .5 vs. .0; P = .04, Wilcoxon rank-sum test) and were significantly older (70.5 vs 53.0 y; P < .0001, Wilcoxon rank-sum test). Moreover, review of the ICD-9 codes in the perioperative period suggests that several of these cases of surgical mortality may have undergone “procedures for creation of esophagogastric sphincteric competence” during the performance of emergency surgery for gastrointestinal hemorrhage (n = 5) or diaphragmatic hernia with obstruction (n = 2). There were 10 cases specifically coded as having surgical or iatrogenic complications, including hemorrhage complicating a procedure (n = 3), and 1 case each of laceration of the esophagus, unspecified accidental surgical laceration, complications of the digestive system, colon injury, complications of the heart with cardiac arrest, respiratory complications, and iatrogenic pulmonary embolism and infarction. There were 3 other cases each of cardiac arrest or respiratory failure, 2 cases each of acute myocardial infarction, shock, and aspiration pneumonia, and 1 case each of cerebrovascular accident and pneumonia. Some patients had more than 1 of these diagnoses.Table 1Demographic Characteristics of the Fundoplication CohortsCharacteristicsOverall cohort (n = 3145)Medication cohortaThe Medication Cohort excluded patients who lacked any of the following: (1) an outpatient visit at least 6 months after surgery. (2) at least 1 outpatient visit during October 1, 1998, to February 28, 2001, or (3) a minimum of 1 year of follow-up evaluation after surgery. (n = 2406)Recent cohortbThe Recent Cohort was composed of patients having surgery between October 1, 1998, and February 28, 2000, allowing for complete pharmacy records of medication use in the perioperative period. (n = 404)Age, y (range)53.8 ± 13 (20–87)53.3 ± 13 (22–86)52.3 ± 12 (23–86)Median follow-up period, y4.55.01.7Male94.4%94.0%89.9%White88.3%88.8%82.9%a The Medication Cohort excluded patients who lacked any of the following: (1) an outpatient visit at least 6 months after surgery. (2) at least 1 outpatient visit during October 1, 1998, to February 28, 2001, or (3) a minimum of 1 year of follow-up evaluation after surgery.b The Recent Cohort was composed of patients having surgery between October 1, 1998, and February 28, 2000, allowing for complete pharmacy records of medication use in the perioperative period. Open table in a new tab Table 2Complications After Antireflux SurgeryComplicationN (%)Months to onset, median (range)Dysphagia609 (19.4)18.2 (.1–121)Dilation202 (6.4)13.8 (.2–116)Redo-fundoplication72 (2.3)15.8 (1.1–103)30-day surgical mortality26 (.8)N/ANOTE. N = 3145. Open table in a new tab Table 3Predictors of Complications After Antireflux Surgery Using Multivariate Regression Models (Hazard Ratio or Odds Ratio [95% Confidence Interval])DilationDysphagiaRedo-fundoplicationSurgical mortalityAge (per year)1.01 (1.00–1.02) P = .0141.01 (1.00–1.02) P = .0007.99 (.98–1.01) P = .381.09 (1.05–1.14) P < .0001White1.27 (.82–1.98) P = .28.88 (.69–1.13) P = .32.89 (.44–1.79) P = .74.66 (.22–1.97) P = .45Male.89 (.51–1.56) P = .69.66 (.49–.90) P = .008.54 (.24–1.18) P = .121.01 (.13–7.66) P = .99Recent surgery (after January 1, 1997).84 (.61–1.17) P = .303.16 (2.60–3.85) P < .00011.44 (.85–2.43) P = .171.35 (.62–2.96) P = .45Log of hospital volume of surgeries1.01 (.87–1.18) P = .86.94 (.85–1.04) P = .261.06 (.78–1.44) P = .71.88 (.55–1.39) P = .57NOTE. Cox proportional hazard models were used for the dilation, dysphagia, and redo-fundoplication models. Logistic regression was used for the surgical mortality model. Open table in a new tab NOTE. N = 3145. NOTE. Cox proportional hazard models were used for the dilation, dysphagia, and redo-fundoplication models. Logistic regression was used for the surgical mortality model. Table 1 shows the baseline characteristics of the medication cohort and the recent medication cohort, for whom prescriptions dispensed in the perioperative period should be complete. Antireflux medications were dispensed commonly to the study cohort (Table 4), with chronic dispensing of antireflux medications to 49.8% of the cohort during the study period, excluding antacids. Many patients received prescriptions from more than 1 category of antireflux medications. Among those patients meeting our criteria for chronic medication use, the median number of days of medication supplied was 210 days for antacids, 450 days for H2RAs, 540 days for PPIs, and 300 days for promotility agents during the study period after October 1, 1998. Analysis of the recent surgical cohort showed that PPI dispensing decreased from 40.1% in the initial 3 months after surgery to 8.2% during the next 3 months, then increased to 14.4% after 6 months (Figure 1). Similar patterns were seen for the other antireflux medications.Table 4Medication Use After Antireflux SurgeryN (%)Any prescription >6 months after surgery Antacid262 (10.9) H2RA885 (36.8) PPI1016 (42.2) Promotility agent352 (14.6)At least 3 prescriptions, including one >6 months after surgery Antacid147 (6.1) H2RA573 (23.8) PPI824 (34.3) Promotility agent221 (9.2) None of the above (excluding antacids)1207 (50.2)NOTE. N = 2406. Open table in a new tab NOTE. N = 2406. Our results indicate that antireflux surgery in the VA medical system is associated with moderate rates of complications and substantial rates of antireflux medication use after a median of approximately 5 years of follow-up. Although there is a pattern of decreased use of these medications in the 3 to 6 months after surgery, approximately 50% of patients are receiving multiple antireflux prescriptions at a median of almost 5 years. Fortunately, significant complications (ie, mortality, redo-surgery) are infrequent in veterans undergoing antireflux surgery, although dysphagia and esophageal dilation are relatively common. We found that dysphagia was coded for 20% of patients, although this is likely an underestimate because the databases are prone to underreport the occurrence of dysphagia, especially before fiscal year 1997 when outpatient ICD-9 codes first were used. Moreover, many patients may have experienced dysphagia that was not severe enough to warrant a clinic visit or was not coded by the examining provider. Dysphagia can range from a minor annoyance with occasional hanging-up of solid food to significant impairment of swallowing. Vakil et al16Vakil N. Shaw M. Kirby R. Clinical effectiveness of laparoscopic fundoplication in a U.S. community.Am J Med. 2003; 114: 1-5Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar found that 27% of patients reported dysphagia when surveyed an average of 20 months after fundoplication. Other investigators have reported temporary postoperative dysphagia in up to 50% of patients 1 week after surgery.17Kamolz T. Bammer T. Pointner R. Predictability of dysphagia after laparoscopic Nissen fundoplication.Am J Gastroenterol. 2000; 95: 408-414Crossref PubMed Google Scholar In a meta-analysis of more than 2400 fundoplications by Perdikis et al,18Perdikis G. Hinder R.A. Lund R.J. et al.Laparoscopic Nissen fundoplication where do we stand?.Surg Laparosc Endosc. 1997; 7: 17-21Crossref PubMed Scopus (210) Google Scholar transient dysphagia was reported in about 20% of patients, with more persistent symptoms in only 6%. Although esophageal dilation reportedly is performed in 11%–23% of patients undergoing antireflux surgery,10Spechler S.J. Lee E. Ahnen D. et al.Long-term outcome of medical and surgical therapies for gastroesophageal reflux disease follow-up of a randomized controlled trial.JAMA. 2001; 285: 2331-2338Crossref PubMed Scopus (851) Google Scholar, 16Vakil N. Shaw M. Kirby R. Clinical effectiveness of laparoscopic fundoplication in a U.S. community.Am J Med. 2003; 114: 1-5Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar, 19Malhi-Chowla N. Gorecki P. Bammer T. et al.Dilation after fundoplication timing, frequency, indications, and outcome.Gastrointest Endosc. 2002; 55: 219-223Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar, 20Khaitan L. Ray W.A. Holzman M.D. et al.Health care utilization after medical and surgical therapy for gastroesophageal reflux disease a population-based study, 1996 to 2000.Arch Surg. 2003; 138: 1356-1361Crossref PubMed Scopus (9) Google Scholar dilation was less common in our cohort (6.4%), for unclear reasons. Only 53 of these patients (1.7%) underwent dilation in the first 3 months after surgery. We found the proportion of patients undergoing a second fundoplication (2.3%) to be similar to the rate of 3.1% in a meta-analysis by Perdikis et al.18Perdikis G. Hinder R.A. Lund R.J. et al.Laparoscopic Nissen fundoplication where do we stand?.Surg Laparosc Endosc. 1997; 7: 17-21Crossref PubMed Scopus (210) Google Scholar Spechler et al,10Spechler S.J. Lee E. Ahnen D. et al.Long-term outcome of medical and surgical therapies for gastroesophageal reflux disease follow-up of a randomized controlled trial.JAMA. 2001; 285: 2331-2338Crossref PubMed Scopus (851) Google Scholar however, reported that 16% of surgical patients had at least 1 additional antireflux surgery over a median of 6.3 years of follow-up. Our 30-day postoperative mortality rate of .83% is slightly higher than that reported by published case series from some specialized esophageal surgeons at tertiary care centers21Bittner H.B. Meyers W.C. Brazer S.R. et al.Laparoscopic Nissen fundoplication operative results and short-term follow-up.Am J Surg. 1994; 167: 193-200Abstract Full Text PDF PubMed Scopus (142) Google Scholar, 22Collard J.M. de Gheldere C.A. De Kock M. et al.Laparoscopic antireflux surgery. What is real progress?.Ann Surg. 1994; 220: 146-154Crossref PubMed Scopus (112) Google Scholar, 23Cadiere G.B. Houben J.J. Bruyns J. et al.Laparoscopic Nissen fundoplication technique and preliminary results.Br J Surg. 1994; 81: 400-403Crossref PubMed Scopus (140) Google Scholar, 24Hinder R.A. Filipi C.J. Wetscher G. et al.Laparoscopic Nissen fundoplication is an effective treatment for gastroesophageal reflux disease.Ann Surg. 1994; 220: 472-483Crossref PubMed Scopus (554) Google Scholar or from Finland.25Rantanen T.K. Salo J.A. Sipponen J.T. Fatal and life-threatening complications in antireflux surgery analysis of 5,502 operations.Br J Surg. 1999; 86: 1573-1577Crossref PubMed Scopus (91) Google Scholar However, data from the Nationwide Inpatient Sample, the largest source of all-payer discharge information in the United States, indicated an unadjusted in-hospital mortality rate of 1.2% in 1990, decreasing to .5% in 1997.6Finlayson S.R. Laycock W.S. Birkmeyer J.D. National trends in utilization and outcomes of antireflux surgery.Surg Endosc. 2003; 17: 864-867Crossref PubMed Scopus (117) Google Scholar Moreover, in a study of 86,411 patients who underwent fundoplication nationwide in the United States between 1992 and 1997, in-hospital death occurred in .8% of patients.26Flum D.R. Koepsell T. Heagerty P. et al.The nationwide frequency of major adverse outcomes in antireflux surgery and the role of surgeon experience, 1992-1997.J Am Coll Surg. 2002; 195: 611-618Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar The odds of death were 5.6 times greater if the fundoplication was one of the first 15 procedures performed by a surgeon when compared with later procedures. Other investigators have found a similar association between surgeon experience and complications.27Trastek V.F. Deschamps C. Allen M.S. et al.Uncut Collis-Nissen fundoplication learning curve and long-term results.Ann Thorac Surg. 1998; 66: 1739-1744Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar, 28Gotley D.C. Smithers B.M. Rhodes M. et al.Laparoscopic Nissen fundoplication—200 consecutive cases.Gut. 1996; 38: 487-491Crossref PubMed Scopus (163) Google Scholar, 29Luostarinen M.E. Isolauri J.O. Surgical experience improves the long-term results of Nissen fundoplication.Scand J Gastroenterol. 1999; 34: 117-120Crossref PubMed Scopus (55) Google Scholar During our study period, a median of 49 antireflux surgeries were performed at the individual VA facilities over 5 years (range, 1–193). We did not find any association between the log of the number of procedures performed and the occurrence of complications. Given that many VA facilities are associated with teaching hospitals, it is possible that many attending surgeons perform this surgery outside of the VA hospital as well, thereby making it difficult to adjust adequately for surgical experience. It is important to keep in mind that differences in the proportion of patients with reported dysphagia, dilation, redo-surgery, and mortality between studies may be explained by differences in duration of follow-up time, patient selection, surgical technique, or in the definition of the outcome. Our finding of common use of acid-suppressive therapy after antireflux surgery confirms the results of other investigators. In a follow-up study of VA patients who were randomized to medical or surgical therapy for the management of GERD, Spechler et al10Spechler S.J. Lee E. Ahnen D. et al.Long-term outcome of medical and surgical therapies for gastroesophageal reflux disease follow-up of a randomized controlled trial.JAMA. 2001; 285: 2331-2338Crossref PubMed Scopus (851) Google Scholar found that although 86% of the surgically treated patients were satisfied or very satisfied with the original surgery, 62% were taking antireflux medications regularly at a median follow-up period of 6.3 years. Overall, GERD symptom scores were similar between medical and surgical patients while taking medication, although medical patients were more symptomatic when medications were withheld. Vakil et al16Vakil N. Shaw M. Kirby R. Clinical effectiveness of laparoscopic fundoplication in a U.S. community.Am J Med. 2003; 114: 1-5Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar reported that 32% of 80 laparoscopic fundoplication patients regularly were taking heartburn medications at an average of 20 months after surgery in a cohort of patients undergoing laparoscopic fundoplication in a managed care organization. Khaitan et al20Khaitan L. Ray W.A. Holzman M.D. et al.Health care utilization after medical and surgical therapy for gastroesophageal reflux disease a population-based study, 1996 to 2000.Arch Surg. 2003; 138: 1356-1361Crossref PubMed Scopus (9) Google Scholar used the Tennessee Medicaid database to evaluate postfundoplication medication use and determined that 74% of patients had received antireflux medications during the fourth year of follow-up. There are several limitations associated with the use of administrative databases for the study of clinical questions. First, there may be inaccuracies in the coding of measured variables and important clinical information may be missing. In the VA system, trained abstractors at each VA medical center perform the ICD-9-CM coding and the data are transferred electronically to the national PTF. By using chart review to verify the PTF data, the reliability of demographic data (κ, ∼.92) and principal diagnoses (κ range, .39–1.0) have been shown to be adequate.30Kashner T. Agreement between administrative files and written medical records a case of the Department of Veterans Affairs.Med Care. 1998; 36: 1324-1336Crossref PubMed Scopus (249) Google Scholar Coding errors do occur, especially concerning non–operating room procedures and secondary diagnoses,31Lloyd S. Rissing J. Physician and coding errors in patient records.JAMA. 1985; 254: 1330-1336Crossref PubMed Scopus (222) Google Scholar although this is more likely to be a problem for clinical diagnoses (eg, dysphagia) than for dilation and redo-fundoplication. Second, although we can determine medication usage indirectly by measuring the type and number of prescriptions dispensed to a patient, we do not know if the patients actually took the medications that were prescribed. Because our primary end point was chronic medication use, only those patients who filled at least 3 prescriptions during the 29 months included in the pharmacy database were considered to be using the medications. This minimizes misclassification of patients, such as those who may have been given a shorter supply of PPI therapy for conditions such as Helicobacter pylori or peptic ulcer disease. Analysis of the number of days of therapy prescribed shows that the average patient classified as using H2RA or PPI therapy received more than a 1-year supply. Moreover, by requiring that at least 1 medication be dispensed at least 6 months after surgery, we minimize the inclusion of patients who mistakenly continue to refill old prescriptions after their surgery. Although some patients received prescriptions for antacids, we cannot determine all over-the-counter medications. However, our study predates the availability of over-the-counter PPI medications. Third, although these medications typically are taken for reflux symptoms, we were unable to determine the indication for these prescriptions from the database. Thus, some patients may take antisecretory or promotility therapy for dyspepsia or other conditions.32Bammer T. Achem S.R. DeVault K.R. et al.Use of acid suppressive medications after laparoscopic antireflux surgery prevalence, clinical indications and causes.Gastroenterology. 2000; 118: A478Abstract Full Text PDF Google Scholar, 33Bammer T. Hinder R.A. Klaus A. et al.Five- to eight-year outcome of the first laparoscopic Nissen fundoplications.J Gastrointest Surg. 2001; 5: 42-48Crossref PubMed Scopus (216) Google Scholar From the perspective of a health care system that weighs the costs of medication vs the costs of surgery, the fact that a significant proportion of surgical patients remain on medical therapy needs to be factored into any cost-effectiveness model, regardless of the indication. Ideally, patients with recurring symptoms should undergo a 24-hour pH probe study to correlate recurring symptoms with objective evidence of acid reflux. Although these studies may have been performed on many patients in the cohort, results of these studies are not available for this analysis. Fourth, this administrative database study does not allow us to ascertain patient satisfaction or symptom control after antireflux surgery. Other studies have shown improvement in quality of life after antireflux surgery.7Duffy J.P. Maggard M. Hiyama D.T. et al.Laparoscopic Nissen fundoplication improves quality of life in patients with atypical symptoms of gastroesophageal reflux.Am Surg. 2003; 69: 833-838PubMed Google Scholar, 8Trus T.L. Laycock W.S. Waring J.P. et al.Improvement in quality of life measures after laparoscopic antireflux surgery.Ann Surg. 1999; 229: 331-336Crossref PubMed Scopus (114) Google Scholar, 34Kamolz T. Granderath P.A. Bammer T. et al.Mid- and long-term quality of life assessments after laparoscopic fundoplication and refundoplication a single unit review of more than 500 antireflux procedures.Dig Liver Dis. 2002; 34: 470-476Abstract Full Text PDF PubMed Scopus (30) Google Scholar, 35Nilsson G. Wenner J. Larsson S. et al.Randomized clinical trial of laparoscopic versus open fundoplication for gastro-oesophageal reflux.Br J Surg. 2004; 91: 552-559Crossref PubMed Scopus (71) Google Scholar, 36Rattner D.W. Measuring improved quality of life after laparoscopic Nissen fundoplication.Surgery. 2000; 127: 258-263Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar Although we have shown significant medication use after surgery, it is conceivable that patients would require additional therapy had they not undergone surgery. In fact, Spechler et al9Spechler S.J. The Department of Veterans Affairs Gastroesophageal Reflux Disease Study GroupComparison of medical and surgical therapy for complicated gastroesophageal reflux disease in veterans.N Engl J Med. 1992; 326: 786-792Crossref PubMed Scopus (519) Google Scholar showed that, when compared with medically treated patients, surgical patients had equal symptom scores on medications, but less symptoms off medications. Moreover, surgical patients had less bodily pain than medical patients on the SF-36 quality-of-life survey. Next, our database is likely to underestimate complications and medication use because we can only ascertain treatment received at VA facilities. Although we excluded patients who did not visit VA facilities during the era of the pharmacy database, assuming that all 268 patients who were not seen were not taking any medications would result in the conclusion that 45% of patients continued to take H2RAs, PPIs, or promotility agents. Also, our cohort may have included some patients who underwent surgery under emergent conditions to manage acute obstruction or hemorrhage. Finally, results from VA facilities may not be generalizable to non-VA populations.22Collard J.M. de Gheldere C.A. De Kock M. et al.Laparoscopic antireflux surgery. What is real progress?.Ann Surg. 1994; 220: 146-154Crossref PubMed Scopus (112) Google Scholar Despite the limitations of this administrative database study, there are unique strengths to this type of analysis. First, unlike many studies from centers of excellence in antireflux surgery, this study reflects the outcomes of patients receiving care from 119 facilities nationwide, including surgeons at various levels of experience. Our ability to link national pharmacy dispensing data to inpatient and outpatient diagnoses and procedures is an important feature of this dataset, allowing for the tracking of patients who receive surgery at 1 facility and are treated for complications or are prescribed medications at another facility. Thus, unlike carefully controlled trials, this effectiveness study describes outcomes that may be more generalizable to the typical patient considering antireflux surgery who receives care from the usual caregivers, rather than from expert providers. Finally, these population-derived data are useful in the refinement of cost-effectiveness studies of surgical vs medical therapy, which recently have found medical management to be more cost effective than surgery.37Arguedas M.R. Heudebert G.R. Klapow J.C. et al.Re-examination of the cost-effectiveness of surgical versus medical therapy in patients with gastroesophageal reflux disease the value of long-term data collection.Am J Gastroenterol. 2004; 99: 1023-1028Crossref PubMed Scopus (37) Google Scholar, 38Dire C.A. Jones M.P. Rulyak S.J. et al.The economics of laparoscopic Nissen fundoplication.Clin Gastroenterol Hepatol. 2003; 1: 328-332Abstract Full Text Full Text PDF PubMed Scopus (12) Google Scholar In summary, our study indicates that patients undergoing antireflux surgery have moderate rates of dysphagia and esophageal dilation, with less frequent redo-fundoplication and surgical mortality rates of approximately .8%. At a median follow-up time of 5 years, nearly half of all patients receive multiple prescriptions for antireflux medications. Therefore, although antireflux surgery has been shown to result in excellent symptomatic control and patient satisfaction in expert centers, patients considering surgery should be counseled fully about the risk for complications and the likelihood of continued antireflux medication use before surgery is performed.
Referência(s)