Effective Clinical Practice
Effective Clinical Practice, May/June 2002
Samuel R.G. Finlayson, MD, MPH, VA Outcomes Group, VA Medical Center, White River Junction, Vt, Department of Surgery, Dartmouth Medical School, Hanover, NH; Kevin T. Stroupe, PhD, George J. Joseph, MS, VA Information Resource Center, Midwest Center for Health Services and Policy Research, Hines, Ill; Elliott S. Fisher, MD, MPH, VA Outcomes Group, VA Medical Center, White River Junction, Vt, Center for the Evaluative Clinical Sciences, Dartmouth Medical School, Hanover, NH
For author affiliations, current addresses, and contributions, see end of text.
Context. In the private sector, the use of surgery to treat gastroesophageal reflux disease has increased substantially since the development of minimally invasive laparoscopic techniques. However, trends in the use of antireflux surgery in the Veterans Affairs (VA) health care system have not been explored.
Objective. To compare secular trends in the use of antireflux surgery in VA hospitals and the private sector.
Data Sources. VA data are from the 1991-1999 medical SAS datasets for inpatient care (commonly known as patient treatment files); private sector data are from the 1991-1997 Nationwide Inpatient Sample and the U.S. census.
Calculations. We compared secular trends in the use of antireflux surgery in the VA and private sector with each group's baseline rate in 1991. For the VA, we calculated annual rates of antireflux surgery among active users of the VA health care system by dividing the number of procedures (based on the appropriate procedure codes from the International Classification of Diseases, ninth revision, clinical modification) by the number of veterans who had at least two hospital or clinic visits in a given year. For the private sector, we calculated true population rates by dividing procedure counts by the total U.S. population.
Results. From 1991 to 1995, the annual rate of antireflux surgery among active users of VA hospitals increased by 64%, then decreased over the next 4 years to almost baseline rates. In contrast, rates of antireflux surgery in the private sector increased 185% from 1991 to 1995, then appeared to reach a plateau thereafter. Among patients undergoing antireflux surgery, those in the VA were less likely than those in the private sector to undergo laparoscopic surgery (29% vs. 65%, respectively, in 1997).
Conclusions. With the development of laparoscopic surgery, rates of antireflux surgery in VA hospitals increased only modestly compared with the private sector and have decreased in recent years. Both patient and provider factors may explain the substantially higher use of this procedure in the private sector.
Population-based studies have shown that rates of surgery for gastroesophageal reflux disease increased substantially in the 1990s.(1-3) Much of this increase has been attributed to the development of a laparoscopic technique for performing antireflux surgery. Because of perceived reductions in operative morbidity and recovery times with this technique,(4-7) the availability of laparoscopic antireflux surgery may have lowered the threshold for surgical intervention for gastroesophageal reflux disease.
Whether a similar phenomenon has occurred in the VA health care system is not known. Use of antireflux surgery in this system may differ from that in the private sector for two reasons. First, because VA patients are on average older and have more comorbid conditions, surgeons may have a higher threshold to operate on them. Second, financial incentives to adopt new surgical technologies may be weaker in the VA. In contrast to their private sector counterparts, VA physicians are salaried. Unlike most private sector hospitals, VA hospitals do not have to compete for patients by offering the latest treatments and technologies.
We examined comprehensive administrative data from the VA health care system to describe how utilization of antireflux surgery changed from 1991 to 1999, the period of greatest growth in the use of laparoscopic techniques. We then compared our findings with the results of an identical query of the Nationwide Inpatient Sample (NIS), a large, federally administered database containing information on approximately 20% of non--VA hospital discharges in the United States.
VA Procedure Rates
Institutional review board approval was granted for the study. We obtained selected records from the VA medical SAS datasets for inpatient care (more commonly known as patient treatment files) for 1991 through 1999.(8) To identify hospitalizations during which an antireflux procedure was performed, we identified all records with International Classification of Diseases, ninth revision, clinical modification (ICD-9-CM) procedure codes 44.65 (esophagogastroplasty) or 44.66 (other procedure for creation of esophagogastric sphincter competence). Because code 44.65 is less specific, we excluded records with this code that did not also include an ICD-9-CM code for gastroesophageal reflux, esophagitis, esophageal ulcer, or hiatal hernia (530.10, 530.11, 530.19, 530.81, 530.20, and 553.3). To increase the homogeneity of the cohort, we excluded patients with a diagnosis code for esophageal cancer (150.0--150.5 and 150.8--150.9), achalasia (530.0), or gangrenous or obstructing incarcerated diaphragmatic hernia (551.3 and 552.3). For similar reasons, we also excluded any cases coded as emergency hospital admissions.
During the study period, no specific ICD-9-CM code for laparoscopic antireflux procedures existed. To estimate the number of antireflux procedures performed laparoscopically, we identified all discharge data that included a code for laparoscopic exploration, laparoscopic lysis of adhesions, or laparoscopic cholecystectomy (54.21, 54.51, and 51.23) or that were associated with a live discharge with a length of stay of 2 or fewer days.
To calculate annual procedure rates in the VA system, we estimated the number of patients actively receiving care within the VA health care system (denominator). The number of VA-enrolled veterans who receive most of their medical care outside this system is substantial and may have changed over time. As a result, simply using the number of enrolled veterans as a denominator for annual rates of antireflux surgery might be misleading. To achieve a more stable denominator, we restricted our analysis to patients in the VA medical SAS datasets who had visited a VA clinic or hospital twice during the year. Thus, the antireflux surgery rates that we calculated for the VA represent rates of surgery for active users of the VA health care system, not rates of surgery for VA-enrolled veterans or rates for the veteran population in general.
Private Sector Procedure Rates
Discharge data from the NIS for the period 1991 through 1997 were obtained from the Healthcare Cost and Utilization Project of the Agency for Healthcare Research and Quality. Further information on the NIS is available elsewhere.(9) The same algorithm used to identify antireflux procedures in the VA dataset was applied to the NIS data. Estimates of national procedure counts were calculated by using hospital sampling weights provided with the NIS data. For the denominator for national procedure rates, we used year-specific adult (age >17 years) population estimates from the U.S. census (true population-based rates).
Because the denominators were dissimilar (patients actively receiving care in the VA vs. total U.S. population), we could not directly compare absolute rates of surgery across the two systems of care. For this reason, we focused instead on secular trends in surgery use relative to each group's baseline rate in 1991.
For both the VA and the NIS cohorts, secular trends in patient sex, patient race, and changes in the proportion of patients undergoing laparoscopic vs. open procedures were tested for significance by performing logistic regression analysis, using year as a continuous variable.(10) We used the Wilcoxon rank-sum test to evaluate secular trends in patient age and length of stay.(11)
Between 1991 and 1999, 2937 patients underwent antireflux surgery in VA hospitals. From 1991 to 1997, approximately 119,700 antireflux procedures were performed in the private sector. Table 1 shows demographic characteristics of the VA and NIS cohorts.
Trends in use of antireflux surgery in the VA differed from those in the private sector (Figure 1). From 1991 to 1995, rates of surgery in the VA increased by 64%, then decreased slowly to almost baseline (1991) rates over the ensuing 4 years. In contrast, rates of antireflux surgery in the private sector increased 185% from 1991 to 1995, then appear to reach a plateau.
The proportion of procedures performed laparoscopically in the VA increased from 0% in 1991 to a peak of 28% in 1999. In contrast, the proportion of procedures performed laparoscopically in the NIS cohort increased from 1% in 1991 to 65% in 1997 (Table 1).
Length of stay in the VA decreased from 12 days in 1991 to 4 days in 1999 (P<0.01). In the NIS cohort, length of stay decreased from a median of 7 days in 1991 to only 2 days in 1997 (P<0.01).
Annual rates of antireflux surgery in VA hospitals increased by 64% in the early 1990s, then decreased to almost baseline (1991) levels by 1999. This pattern differs from utilization in the private sector, in which rates of surgery increased much more sharply in the mid-1990s. The proportion of patients undergoing laparoscopic antireflux surgery increased in both the private sector and the VA, but did so more slowly in the latter group. By 1999, use of laparoscopy in the VA had only reached levels observed in the private sector before 1993.
Three previous population-based studies of the use of antireflux surgery have been done. A national study from Finland showed a 76% increase in annual surgery rates between 1988 and 1993.2 In a study from Canada,1 investigators found only a 14% increase from 1992 to 1996. A study from the U.S. assessed trends in antireflux surgery in three New England states, where surgery rates increased 129% between 1993 and 1998.3 Although we focused on a slightly different period, our findings are generally consistent with those of the Finnish and U.S. studies.
Comparisons of utilization of antireflux surgery in VA patients and non-VA patients have not been published; however, this comparison has been made for cholecystectomy, another procedure for which laparoscopy has become the preferred surgical approach.(12) That study showed a 13% increase in use of cholecystectomy in the VA during a period when the proportion of laparoscopic procedures increased from 0% to 52%. According to several other population-based studies, cholecystectomy rates increased faster in the private sector.(13-17)
The increase in utilization of antireflux surgery nationally is probably attributable to lower thresholds for surgical intervention since the development of laparoscopic techniques for this procedure. Laparoscopic surgery may be more attractive to patients because it results in less postoperative pain and a quicker return to full activity. Physicians may also be more likely to recommend surgery because of favorable reports on the efficacy of antireflux surgery published in the 1990s.(18)
However, other factors may have contributed to the sharp increase in use of surgery to treat acid reflux. This very common clinical condition has become medicalized: "Heartburn" has become "gastroesophageal reflux disease." Direct-to-consumer advertising of proton-pump inhibitors recommends that patients with heartburn discuss their condition with their physicians. Thus, the number of patients who seek medical attention and receive a diagnosis of gastrointestinal reflux disease has undoubtedly increased. Even if laparoscopic surgery did not lower the threshold to operate on individual patients, a larger reservoir of eligible patients would have led to more patients being referred for surgery.
Although we have documented growth in the use of antireflux surgery nationally, our analysis does not explain why utilization of antireflux surgery in the VA did not keep pace with utilization in the private sector. However, this observation might be at least partially explained by several important differences between the VA and the private sector. First, clinical characteristics (such as age and comorbid conditions) of patients in the VA differ from those of patients in the private sector. The threshold to operate may be higher in patients with greater perceived operative risk. Second, many enrolled VA patients also have access to health care in the private sector and may choose to use it selectively based on specific health care needs. Patients who otherwise receive primary care within the VA system may move "out of system" for advanced laparoscopic surgery, causing us to underestimate use of this procedure in veterans. Third, financial incentives to perform laparoscopic antireflux surgery may have contributed to more rapid growth in utilization in the private sector. To the extent that financial incentives are much weaker in the VA system, growth of antireflux surgery in the VA would be expected to be slower. Finally, institutional commitment to providing advanced laparoscopic surgery may be weaker in the VA. In an effort to increase market share, private hospital administrators may have stronger incentives to establish services that are perceived as providing the most advanced care available.
Our study has several limitations. Our ability to identify laparoscopic procedures was hampered by the absence of specific ICD-9-CM codes for laparoscopic antireflux surgery. We estimated the proportion of procedures performed laparoscopically by the presence of other procedure codes specific for laparoscopy and by length of stay. We suspect that this technique led to conservative estimates; however, other investigators have validated the use of a similar code-based protocol to study cholecystectomy before the appearance of a specific code for laparoscopic cholecystectomy.(16)
Second, we could not directly compare absolute rates of antireflux surgery in the VA and the private sector. In our calculation of annual procedure rates in the VA, we used as a denominator the number of unique patients in the inpatient and outpatient medical SAS datasets who had visited a VA clinic or hospital twice during the year. Because we applied this denominator consistently, our analysis should accurately reflect relative changes in surgery rates over time.
Finally, private sector data beyond 1997 were lacking. We do not know whether rates of antireflux surgery in the private sector increased or decreased (as in the VA) in the late 1990s. However, a population-based study of antireflux surgery in New England demonstrated continued growth in annual rates of antireflux surgery after 1997.
In conclusion, use of antireflux surgery has increased since the development of laparoscopy, but the increase has been considerably slower in VA hospitals than in the private sector. This disparity between the VA and the private sector is no doubt partially attributable to differences in patient characteristics and provider incentives. However, the disparity also underscores the considerable uncertainty about optimal treatment of gastroesophageal reflux disease. We could not determine whether antireflux surgery is underutilized in the VA or overutilized in the private sector. However, our findings emphasize the need for more research to clarify the role of surgery in the treatment of gastroesophageal reflux disease.
1. McMahon RL, Mercer CD. National trends in gastroesophageal reflux surgery. Can J Surg. 2000;43:48-52.
2. Viljakka M, Luostarinen M, Isolauri J. Incidence of antireflux surgery in Finland 1988-1993. Scand J Gastroenterol. 1997;32:415-8.
3. Finlayson SR, Birkmeyer JD, Laycock WS. Trends in surgery for gastroesophageal reflux disease: the effect of laparoscopic surgery on utilization. Surgery. [In press].
4. Laine S, Rantala A, Gullichsen R, Ovaska J. Laparoscopic vs conventional Nissen fundoplication: a prospective randomized study. Surg Endosc. 1997;11:441-4.
5. Pelgrims N, Closset J, Sperduto N, Gelin M, Houben JJ. What did the laparoscopic Nissen approach of the gastro-oesophageal reflux really change for the patients 8 years later? Acta Chir Belg. 2001;101:68-72.
6. Eshraghi N, Farahmand M, Soot SJ, Rand-Luby L, Deveney CW, Sheppard BC. Comparison of outcomes of open versus laparoscopic Nissen fundoplication performed in a single practice. Am J Surg. 1998;175:371-4.
7. Richards KF, Fisher KS, Flores JH, Christensen BJ. Laparoscopic Nissen fundoplication: cost, morbidity, and outcome compared with open surgery. Surg Laparosc Endosc Percutan Tech. 1996;6:140-3.
8. The Medical SAS(r) Inpatient Datasets-FY2000: A VIReC Resource Guide. Edward J. Hines, Jr. VA Hospital, Hines, IL: Veterans Affairs Information Resource Center; 2001.
9. Steiner C, Elixhauser A, Schnaier J. The healthcare cost and utilization project: a family of databases. Eff Clin Pract. 2002;3:XX-XX.
10. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York: Wiley; 1989.
11. Cuzick J. A Wilcoxon-type test for trend. Stat Med. 1985;4:87-90.
12. Ferreira MR, Bennett RL, Gilman SC, Mathewson S, Bennett CL. Diffusion of laparoscopic cholecystectomy in the Veterans Affairs health care system, 1991-1995. Eff Clin Pract. 1999;2:49-55.
13. Nenner RP, Imperato PJ, Rosenberg C, Ronberg E. Increased cholecystectomy rates among Medicare patients after the introduction of laparoscopic cholecystectomy. J Commun Health. 1994;19:409-15.
14. Legorreta AP, Silber JH, Costantino GN, Kobylinski RW, Zatz SL. Increased cholecystectomy rate after the introduction of laparoscopic cholecystectomy. JAMA. 1993;270:1429-32.
15. Orlando R, Russell JC, Lynch J, Mattie A. Laparoscopic cholecystectomy. A statewide experience. The Connecticut Laparoscopic Cholecystectomy Registry. Arch Surg. 1993;128:494-9.
16. Steiner CA, Bass EB, Talamini MA, Pitt HA, Steinberg EP. Surgical rates and operative mortality for open and laparoscopic cholecystectomy in Maryland. N Engl J Med. 1994;330:403-8.
17. Escarce JJ, Chen W, Schwartz JS. Falling cholecystectomy thresholds since the introduction of laparoscopic cholecystectomy. JAMA. 1995;273:1581-5.
18. Spechler JS. Comparison of medical and surgical therapy for complicated gastroesophageal reflux disease in veterans. N Engl J Med. 1992;326:786-92.
The views expressed herein do not necessarily represent the views of the U.S. Department of Veterans Affairs or the U.S. Government.
Dr. Finlayson is supported by a Career Development Award from the VA Health Services Research and Development Program.