Medical or Surgical Therapy for Erosive Reflux Esophagitis

Abstract

Objective

To compare the cost and utility of healing and maintenance regimens of omeprazole and laparoscopic Nissen fundoplication (LNF) in the framework of the Canadian medical system.

Summary Background Data

Medical therapy with proton pump inhibitors for endoscopically proven reflux esophagitis is a safe and effective treatment option. Of late, the surgical treatment of choice for this disease has become LNF.

Methods

The authors’ base case was a 45-year-old man with erosive reflux esophagitis refractory to H₂-blockers. A cost-utility analysis was performed comparing the two strategies. A two-stage Markov model (healing and maintenance phases) was used to estimate costs and utilities with a time horizon of 5 years. Discounted direct costs were estimated from the perspective of a provincial health ministry, and discounted qual-ity-of-life estimates were derived from the medical literature. Sensitivity analyses were performed to test the robustness of the model to the authors’ assumptions and to determine thresholds. A Monte Carlo simulation of 10,000 patients was used to estimate variances and 95% interpercentile ranges.

Results

For the 5-year period studied, LNF was less expensive than omeprazole ($3519.89 vs. $5464.87 per patient) and became the more cost-effective option at 3.3 years of follow-up. The authors found that 20 mg/day omeprazole would have to cost less than $38.60 per month before medical therapy became cost effective; conversely, the cost of LNF would have to be more than $5,273.70 or the length of stay more than 4.2 days for medical therapy to be cost effective. Estimates of quality-adjusted life-years did not differ significantly between the two treatment options, and the incremental cost for medical therapy was $129,665 per quality-adjusted life-years gained.

Conclusions

For patients with severe esophagitis, LNF is a cost-effective alternative to long-term maintenance therapy with proton pump inhibitors.

Symptoms of heartburn affect 7% of the North American adult population daily, 14% weekly, and 36% monthly. ¹ Three percent of heartburn sufferers have severe symptoms. ² Epidemiologic studies in ambulatory subjects have demonstrated that reflux is not limited by geography or culture. Seventy-three percent of subjects in Minnesota had moderate heartburn or regurgitation at least weekly. ³ In Finland, 54% of elderly women and 66% of the men had heartburn monthly. ⁴ In Sweden, 25% of the respondents to a questionnaire admitted to frequent heartburn. ⁵ These data suggest that symptomatic heartburn is one of the most common human ailments, despite apparent variations in its rate of occurrence. Further, it has recently been shown that there is a strong association between symptomatic reflux and esophageal adenocarcinoma (odds ratio 7.7). ⁶

Distal esophageal erosions visualized on endoscopy indicate gastroesophageal reflux disease (GERD). While the symptom of heartburn is common, endoscopic changes are found in a minority of individuals. However, the finding of endoscopic GERD is significant in that it is highly predictive of poor healing rates with H₂-blockers or prokinetic drugs. Nevertheless, proton pump inhibitors (PPIs) appear to offer significant therapeutic advantage in healing the subgroup of individuals with grade II–IV esophagitis, although healing success is dependent on the grade of esophagitis at presentation, with grade IV esophagitis having a much higher medical failure rate. ^7,8

The most frustrating aspect of GERD treatment is the high relapse rate after successful medical healing. More than 80% of patients with erosive (grade II or higher) esophagitis will relapse within 6 months, with 50% of the relapses occurring in the first month. ^9,10 This observation has led some to conclude that maintenance therapy is necessary for all individuals with endoscopically proven reflux disease. Different classes of drugs have been studied in this context. Maintenance trials with prokinetic agents and H₂-blockers have demonstrated modest efficacy in the severe esophagitis group, suggesting that the so-called “stepdown” approach may not be possible after healing is achieved with PPIs. An expert consensus has suggested that the drug successful in healing should be the drug chosen as maintenance. ¹¹

Before the advent of potent antisecretory drugs, surgery was a keystone in the long-term management of GERD. Open Nissen fundoplication was introduced by Rudolph Nissen in 1956 and has since gained respect as an effective alternative to medical maintenance therapy. Long-term follow-up data suggest that up to 90% of patients are symptom-free and require no maintenance medications after 20 years of observation. ^12,13 While there are many established surgical approaches, they can all be categorized as either complete (360°) or partial (180°) wraps. The Nissen procedure is a complete fundoplication, whereas the Belsey Mark IV, Dor, and Toupet repairs are partial and have generally not been as successful in restoring lower esophageal sphincter competence. ^14,15

Since 1991, the laparoscopic Nissen fundoplication (LNF) has shown physiologic results similar to those of the open technique. ^16,17 Ambulatory monitoring of esophageal pH in pre- and post-LNF patients has demonstrated a 10-fold reduction in acid exposure time in the distal esophagus. ^18,19 Laparoscopic surgery was also found to normalize the 24-hour pH study in 91% of subjects. ²⁰ Physiologic measurements of lower esophageal sphincter pressure demonstrated a mean increase of 8.3 to 16.2 mm Hg after LNF. ^21–24 Thus, the effectiveness of LNF is similar to that of its open counterpart with over 3 years of follow-up. ²⁵ Preliminary results (n = 310) from a European multicenter randomized trial showed no differences in the incidence of either endoscopic or symptomatic relapse between medical and laparoscopic therapy, with only 6.7% relapsing in the surgical arm over 5 years. ²⁶ Quality of life was similar between the two groups at up to 5 years of follow-up. ^27,28 For those individuals requiring long-term maintenance therapy for GERD, LNF is becoming an increasingly viable option, with the perceived advantages of decreased cost and avoidance of long-term drug side effects. ²⁹ In this study, we sought to compare the costs and clinical consequences of two treatment options for erosive reflux esophagitis: LNF versus maintenance medical therapy with omeprazole.

METHODS

Markov Model

A two-stage Markov model was constructed using the DATA (TreeAge, Boston, MA) software package. ³⁰ Our base case was a 45-year-old man with endoscopically proven grade II–IV erosive esophagitis, refractory to H₂-blockers. We chose a time horizon of 5 years because of currently available data on relapse rates. ³¹ Our simulation considered two treatment options: medical therapy with omeprazole for endoscopic healing and maintenance versus surgery using LNF.

The Markov model of prognosis is one example of an alternative to standard decision analytical techniques wherein transitions between health and various disease states are modeled over time. ³² Instead of assuming, for example, a one-time probability of therapeutic failure as in a standard decision analysis, the model instead estimates a certain probability for failure during each time cycle.

One intrinsic limitation to the Markov model is the lack of memory. ³² The model assumes that the previous experience of the person in the model does not influence the probabilities of subsequent events (i.e., transition probabilities are independent of one another). In our analysis, a certain degree of memory was required for the medical arm of the model, as the maintenance dose was dependent on the healing dose required. Thus, a two-stage model was constructed, which included the creation of five separate Markov chains stemming from the five regimens required for successful healing (described below). There is evidence that the initial grade of esophagitis does not influence outcome significantly after LNF, and so modeling of memory for the surgical arm was not performed. ³³

Figure 1 illustrates the mathematical relationships of the health states used in our Markov model. Transitions were allowed at the end of each 3-month cycle. Rates derived from the medical literature were converted to transition probabilities according to previously published methods (using the formula P(t) = 1 - e^-rt, where P(t) is the probability of an event during the time cycle, r is the rate, and t is the duration of the cycle expressed in the same time units used in the rate). ³⁰ Rates were varied using interpolation of an approximated hazard function derived from the literature for transitions for which a constant hazards assumption was not appropriate over the 5-year time period (symptomatic relapse and transition between dosages of omeprazole, postoperative dysphagia). ^25,31 Age- and gender-matched population-based mortality rates (life expectancy 31.8 years and 37.0 years for 45-year-old men and women, respectively) came from published Alberta vital statistics. ³⁴ At the completion of each cycle, costs and utilities were accrued.

Figure 1. Markov states and possible transitions modeled for the maintenance phases of the two treatment options: (A) omeprazole and (B) laparoscopic Nissen fundoplication.

Utilities

Utilities were used as an estimation of quality of life, ranging from an asymptomatic cycle, which accrues 0.25 quality-adjusted life-years (QALYs; i.e., three “healthy” months), and death, which is assigned a utility of 0. Disutilities, which were subtracted from this ideal utility, were assigned to disease states such as recurrent reflux symptoms or postoperative recovery (Table 1). A recent Canadian study used standard gamble and time trade-off techniques in symptomatic reflux patients to estimate the disutility associated with chronic GERD. ³⁵ The investigators found that reflux symptoms were associated with a utility of 0.90 to 0.97, depending on the technique employed. Our model used the mean of these two figures. This disutility was applied for 1 month for each patient who relapsed during a cycle. For those individuals in the medical arm, the disutility associated with chronic omeprazole ingestion was assumed to be similar to that of other chronic medical therapy, such as lipid-lowering agents (0.01). ³⁶ Surgery was assumed to have a disutility of 0.5 for the immediate 2-week postoperative period. For those surgical patients who required dilatation or foreign body removal, the assigned disutility was 0.20 for 1 week. All utilities were discounted at 3% per annum in the maintenance phase.

Table 1. PARAMETER ESTIMATES FOR MODEL

LNF, laparoscopic Nissen fundoplication.

Costs

For the cost analysis, we took the perspective of a Canadian provincial (Alberta) health ministry. Discounted direct costs (in Canadian dollars) were estimated from local costs incurred by the hospital, drug formulary, or provincial health ministry (Table 2). Selected costs were estimated from charges billed to nonresidents. The cost of LNF was estimated from a local costing study carried out at the Grey Nuns Hospital in Edmonton, Canada. Mean length of stay (1.2 days), consumables, and other hospital costs for LNF were calculated by reviewing all such surgeries performed during a 3-month period. Ranges in hospital stay reported in the recent literature were considered in sensitivity analyses. Per-diem costs were estimated from charges billed to non-Alberta residents (Grey Nuns Hospital). Physicians’ fees were derived from the fee schedule of the Alberta Health Care Insurance Plan. For the purposes of the analysis, the financial cost of death was assumed to be zero. All costs were discounted at 3% per annum in the maintenance phase.

Table 2. COST ESTIMATES FOR MODEL

LNF, laparoscopic Nissen fundoplication; AHCIP, Alberta Health Care Insurance Plan.

* Includes costs of office visit, endoscopy, manometry, consumables, nursing, surgeon, anesthetist, and a 1.2-day length of stay (range 1–7 days).

† Includes costs of office visit, endoscopy, anesthetist, surgeon, nursing, consumables, and a 7-day hospital stay (range 4–14 days).

Medical Therapy Arm

The medical arm consisted of an initial healing phase followed by maintenance therapy, both using the PPI omeprazole. In the healing phase, patients were assigned to one of five treatment arms, each one representing the different dose and/or duration of therapy required to accomplish successful endoscopic healing. We based the proportion of patients assigned to each of the five healing regimens primarily on probabilities derived from the work of Klinkenberg-Knol et al, who published detailed healing and maintenance data for patients with refractory erosive (grade II–IV) esophagitis, using endoscopic criteria for healing success and relapse (see Table 1). ³¹ In this cohort, omeprazole at 40 mg daily healed 64% of patients after 1 month, with an additional 18% healing after 2 months and 13% after 3 months. Of the remainder, 4% healed after 4 months, and 1% required 4 months of therapy plus an additional month of omeprazole at 60 mg/day. These observations have been confirmed by other studies. ^37–42

After healing, patients in the medical arm entered the maintenance phase at omeprazole 20 mg/day. Patients requiring more than 4 months of therapy or more than 60 mg/day omeprazole to achieve healing received 40 mg/day omeprazole in the maintenance phase. If patients relapsed, their maintenance dose was escalated by 20-mg/day increments to a maximum of 60 mg/day. Probabilities for relapse were derived from the data of Klinkenberg-Knol et al. ³¹ The following daily doses were required to maintain remission after 1 year of follow-up: 20 mg for 65%, 40 mg for 32%, and 60 mg for 3% of patients. Other investigators have observed similar relapse rates. ^43–47 We assumed that individuals in the maintenance group requiring 60 mg/day omeprazole would opt for LNF at a rate of 20% annually. An initial upper endoscopy was required not only for the medical healing phase but also if the patient required 60 mg omeprazole at any time during the analysis. For all maintenance arms, physician visits were required only for symptomatic patients, with a maximum frequency of every 3 months.

Surgical Therapy

In the surgical arm of the model, the healing phase was considered to be the surgery itself. If one survived the surgery, one entered the Markov model for the maintenance phase. In a study published in 1997, Perdikis et al reviewed the current LNF literature, including 2,453 patients and 24 studies, with up to 3 years (median not available) of follow-up. ²⁵ Table 3 summarizes the literature from 1996 until the present for comparison. In Perdikis et al’s study, the perioperative mortality rate for LNF was found to be 0.2%. It appears from review of more recent literature that even this low rate may be an overestimation. 17,23,48–64 Urgent laparotomy was required for complications (perforation, bleeding, perforation, pneumothorax, splenectomy) in 1.5%, while 5.8% required conversion to an open approach (for reasons of exposure, perforation, bleeding, CO₂ retention, or pneumothorax). In comparison, more recent publications report an average conversion rate of 3.3%. 17,23,48–64 During follow-up, dilatation was required for 3.5% of patients due to dysphagia or gas-bloat type symptoms, and 0.5% required endoscopy for food impaction. Intermediate to long-term dysphagia rates in more recent reports are not markedly different, with a weighted mean of 4.8%. 17,23,48–64 Repeat surgery was required in 1.6%, either for dysphagia/gas-bloat or for recurrent reflux symptoms. In the review by Perdikis et al, ²⁵ 3.4% of patients continued to require antisecretory medications for persisting reflux symptoms.

Table 3. SUMMARY OF THE LITERATURE 1996–2000

Based on the above data, we assumed that among patients who developed recurrent symptoms following LNF, one third would elect to have surgery redone (via the open approach), and the remainder would continue on 40 mg/day omeprazole for the duration of the study. Although four small studies have now shown that laparoscopic re-do procedures may be feasible, this option was not modeled, as larger studies were felt to be needed in this area. ^65–68 For the open approach, a mortality rate of 1.4% was used. ^69–72 Repeat surgical failures were maintained on 40 mg/day omeprazole. If 40 mg/day failed, the dose was increased to 60 mg/day (“Fail Med-Surgery” in Fig. 1).

In the surgical arm, all patients underwent endoscopy not only at presentation but also if 60 mg/day omeprazole was required at any time during the analysis, as in the medical arm. Endoscopy was also carried out if the patient required dilatation or foreign body removal, using the rate estimated by Perdikis et al ²⁵ (see Table 2). A rate of dilatation, after foreign body removal, of 70% was used. ⁷³

Sensitivity Analysis and Monte Carlo Simulation

Sensitivity analyses were performed to test the robustness of the model with respect to the assumed parameter values. Clinically relevant ranges were tested using available literature (see Table 1).

In a Monte Carlo simulation, each patient passes through the model from beginning to end (5 years), with transitions at each cycle decided by a random number generator and the probabilities associated with that transition. ³² The purpose of the simulation is to arrive at an estimate of the precision of the outcomes of interest (costs and utilities). For each arm, 10,000 patients were simulated, yielding estimates of the mean costs and utilities for each strategy as well as a standard deviation and 95% interpercentile ranges for each parameter.

RESULTS

In our base case analysis, LNF was less expensive than omeprazole. Over the 5-year time horizon of the model, the per-patient cost of LNF was $3,519.89, compared to $5,464.87 for medical therapy. This represents an incremental cost of $1,944.98 per patient for medical therapy. QALY estimates were not significantly different among treatment arms: 4.335 and 4.350 for LNF and medical therapy respectively, which equates to a difference of less than 1 “healthy day” over 5 years (Table 4). This corresponds to an incremental cost for medical therapy of $129,665 per QALY gained. We performed sensitivity analyses to test the robustness of the baseline assumptions and to determine thresholds (Figs. 2–4). One-way sensitivity analysis demonstrated that the model’s conclusions were most dependent on the values of three variables: the cost of medical therapy, the cost of surgery, and time.

Table 4. RESULTS OF MODEL

Figure 2. A one-way sensitivity analysis comparing the two treatment options as they vary with the cost of omeprazole (medical therapy: dashed line; surgery: solid line). The threshold (point at which both therapies have equal expected costs) is $38.60/mo. The vertical dotted line represents the cost used in the model ($68.65).

Figure 3. A one-way sensitivity analysis comparing the two treatment options as they vary with the cost of surgery (medical therapy: dashed line; surgery: solid line). The threshold is $5,296.40. The vertical dotted line represents the cost used in the model ($3,091).

Figure 4. A one-way sensitivity analysis comparing costs of the two treatment options as they vary with time (medical therapy: dashed line; surgery: solid line). The threshold is 3.1 years or 12.2 cycles. At 5 years, surgery is the less expensive option.

The results of the Monte Carlo simulation based on 10,000 patients are summarized in Table 5 and Figure 5. Using the Monte Carlo simulation, the 95% interpercentile range for the cost of the medical therapy strategy was $1,912 to $8,628, and the distribution was trimodal, with the largest mode between $4,000 and $5,000, a second smaller one between $7,000 and $8,000, and the smallest between $1,000 and $2,000. This likely reflects the three maintenance dosing possibilities modeled (20, 40, or 60 mg/day omeprazole). which over time have diverging cost trends. The corresponding interval for the surgical arm was $3,091 to $9,548; this distribution was skewed toward lower values, with outliers (a second mode in fact) corresponding to those patients who suffered complications, at $9,000 to $10,000. For utilities, the 95% interpercentile range for medical therapy was 1.48 to 4.73 QALYs; for the surgical arm, it was 1.23 to 4.73 QALYs.

Table 5. RESULTS OF THE MONTE CARLO SIMULATION

Quality-adjusted life-years.

Figure 5. This histogram displays the distribution of possible 5-year costs, using the Monte Carlo simulation of 10,000 hypothetical patients, of the two arms of the model.

The intervals resulting from Monte Carlo simulations must be interpreted with caution as they are a result of a mathematical simulation taking into account the variables and ranges included in the model, for which the distributions are not well known, and may not accurately represent the true variability in the general population. However, based on the results of this analysis, surgical therapy appeared less costly than medical therapy. There was no clinically important difference in quality of life, estimated by QALYs.

Figure 2 depicts the results of a one-way sensitivity analysis on the cost of medical therapy. The monthly cost of omeprazole would have to be reduced to less than $38.60 before medical therapy would become the cheaper option, and to $38.80 before it became cost effective. Likewise, our analysis of the surgical arm suggested that the cost of LNF would have to rise higher than $5,296.40 before the medical option would be the less expensive strategy and $5,273.70 before it became cost effective (see Fig. 3). Because the predominant cost driver of surgical procedures in Canada is the length of hospital stay, we performed a separate one-way analysis on this variable. As the length of stay rose beyond 4.2 days, or the cost of one day in the hospital rose above $1,857, medical therapy became the superior option. A one-way analysis using time as the dependent variable demonstrated that the initial costs of the surgical procedure were recouped after 3.1 years when compared to the medical arm (see Fig. 4), and the cost-effectiveness ratio began to become superior after 3.3 years.

Figure 6 presents a two-way sensitivity analysis illustrating simultaneously the relationship between time and monthly medical costs. This allows one to plot local drug costs and expected maintenance time. For example, if the drug costs were expected to be $40/month and 2 years of medical maintenance therapy were anticipated, medical therapy would then be less expensive than LNF. Other variables, including the cost of endoscopy, office visits, and drug dispensing fees, appeared to affect both strategies equally and therefore did not influence the conclusions of the model.

Figure 6. A two-way sensitivity analysis, simultaneously examining the effects of omeprazole costs and anticipated duration of medical therapy on overall costs of the two strategies. The shaded area under the line highlights the coordinates that correspond to conditions where medical therapy would be the less expensive option. Coordinates falling above the line correspond to conditions for which surgery is the less expensive option.

As LNF is a relatively new therapeutic intervention, the long-term efficacy has yet to be firmly established. In our base case analysis, we assumed a failure rate of 5.5% over 5 years (failure defined as the return of reflux symptoms requiring the use of maintenance omeprazole therapy at any dose). However, several other reports have suggested that the failure rate may be as high as 5% to 10% per year (see Table 1). ^69,74,75 Therefore, our sensitivity analysis used a maximal failure rate of up to 10% per year. At this rate, the 5-year costs for LNF would be $4,030.16. Nevertheless, the annual failure rate for LNF would have to rise above 37% per year before medical therapy became the less expensive option. However, as one would expect, at this failure rate, the utilities for LNF drop approximately linearly to 4.24 QALY. As expected, the utilities associated with quality of life were most sensitive to the long-term effects of postoperative symptoms and daily medication use. A rise above 3% in the annual LNF failure rate would cause the quality of life utilities to become correspondingly less favorable.

DISCUSSION

In this study we sought to determine the costs and consequences of long-term medical or surgical therapy for the subgroup of GERD patients with erosive (grade II or higher) reflux esophagitis. Our analysis illustrates that LNF is a cost-effective option for middle-aged patients with erosive esophagitis when the expected time of medical maintenance therapy is more than 3.1 years.

However, the results of our analysis differ significantly from those of a previous American study on this issue. Heudebert et al ⁷⁶ found that while both strategies were similarly effective, omeprazole was significantly less expensive than LNF. While there are several important differences in methodology, the major distinction is the costs cited in the American model. Costs in Heudebert et al’s study were strikingly higher for all interventions except drug therapy. For example, the cost of LNF was 7,500 U.S. dollars versus 3,091 Canadian dollars in our model. In comparison, most other published estimates of LNF costs have been less than that used in Heudebert et al’s model. The Finnish hospital costs in a randomized prospective study comparing open and laparoscopic fundoplication were recently reported as $2,981. ⁷⁷ A second study comparing open and laparoscopic surgery costs in Sweden showed that the direct costs of LNF were 27,693 SEK. ⁷⁸ A third American estimate of hospital costs of LNF arrived at the figure of $6,000, which was about half that estimated for the open approach. ⁷⁹ Lastly, one of the most recent American cost estimates, using an outpatient surgery program, was $4,588 for hospital costs. Also, the cost of gastroscopy was at least four times higher ($1,105 U.S. dollars) in Heudebert et al’s study than in our analysis. While these differences reflect international variations in the cost of goods and services, they also indicate that medical economic models are not necessarily portable across political boundaries.

Two other studies have looked at quality of life after LNF. One examined this in terms of improvement in symptoms and reduction of medication use and found that by these measures, quality of life improved. ⁸⁰ A second study, examining “psychological well-being” and the Gastrointestinal Symptoms Rating Scale, found that quality of life of patients was restored to normal with surgery by 12 months. ⁸¹

A problem with previous analyses of this issue has been the assumption that the LNF failure rate is the major long-term adverse consequence of the intervention. However, as experience broadens, it is now increasingly recognized that some individuals will experience intermittent dysphagia and food impaction. In fact, a prematurely ended randomized trial comparing open and laparoscopic surgery has suggested a higher rate of dysphagia with laparoscopy. ⁴⁸ Although the “gas-bloat” syndrome can also occur, these symptoms tend to abate within 3 to 6 months of surgery, can usually be treated conservatively and without dilatation, and will, therefore, not have as great an impact on the costs or utilities. ^58,82,83 Incorporating factors such as ongoing adverse symptoms will bias any model against a surgical intervention but needs to be considered in future studies.

While the conclusions of our analysis are robust within the framework and constraints of the model, several issues should lend caution to the interpretation of the results. First, only one of the currently available PPIs (omeprazole) was modeled. While few head-to-head trials of lansoprazole and pantoprazole with omeprazole have been published, the available evidence suggests that all three PPIs have similar healing and maintenance efficacy in GERD. ^84–91 While the newer PPIs have a slightly lower cost, a reduction of at least 50% is required before medical therapy is more cost effective in our model.

A second issue is the potential long-term complications of PPI therapy, which could not be modeled reliably simply because of lack of data. Until recently, carcinoid tumors in the setting of chronic PPI use had not been documented in humans, despite concerns raised by animal studies. However, a Japanese report may represent the first such human complication. ⁹² Although gastrin levels can occasionally become very high, they are not measured routinely as the clinical consequences of hypergastrinemia in this setting are unclear. ⁹³ Fundic gland polyps may occur but are thought to have a benign prognosis and require neither surveillance nor treatment. Atrophic gastritis and argyrophil cell hyperplasia occurred in approximately 20% of patients in the 5-year follow-up study by Kuipers et al. ⁹⁴ Most of the atrophy development was seen in Helicobacter pylori-positive patients, and none of these changes were seen in patients who had undergone fundoplication. Klinkenberg-Knol et al ⁹⁵ have recently shown, in a study with a mean follow-up of 6.5 years, that atrophy occurs in 4.7% and 0.7% of H. pylori-positive and -negative patients, respectively, and that these rates were related to patient age and severity of gastritis at the beginning of therapy. No dysplasia or neoplasm was seen in over 1,000 patient-years of observation. ⁹⁵ Although there remains a concern that H. pylori eradication may predispose to worsening of reflux symptoms, symptomatic relapse was not found to depend on H. pylori status in the large study by Klinkenberg-Knol et al. ⁹⁵ These issues, although important and concerning, could not be addressed in this analysis as their impact on cost and clinical outcome is currently unknown.

Third, esophageal complications of reflux, such as the development of peptic stricture while on medical therapy, are not modeled mainly because rates of occurrence of these outcomes are not accurately known. The dilatations modeled into the surgical group were for complications of the surgical procedure and not for reflux-induced strictures. As well, the outcome of Barrett’s esophagus could not be accurately modeled. While surgical therapy may be more effective than medical therapy in preventing complications such as Barrett’s esophagus, no long-term trials directly comparing the effectiveness of these therapeutic approaches have yet been published. ⁹⁶ A recent study demonstrated a decrease in the frequency of adenocarcinoma in those undergoing surgery in a Veterans Administration multicenter randomized trial (n = 248). ²⁸ As well, the issue of squamous epithelial regrowth over underlying Barrett’s in the patients with apparently successful regression renders even the existing literature regarding success rates too unreliable to use in our model. ⁹⁷

A fourth limitation of our model is the validity of the available data on long-term LNF success. Unfortunately, a well-known bias in the literature leans toward the reporting of favorable results. We acknowledge that the literature likely reflects the experience of the best centers and may not apply specifically to every surgeon. A significant bias in the opposite direction, however, is the inclusion of results from early in the LNF era. Because of the long learning curve, ^52,98,99 these results are likely slightly worse than current practice, since complication rates appear to continue to fall even after 100 procedures. ^52,98,99 Surgical costs may vary slightly from center to center just as length of stay may vary from surgeon to surgeon. However, despite these possible discrepancies, sensitivity analyses verified that our conclusions were reliable up to five times the reported failure rate, up to $2,000 above our surgical costs, and up to a mean length of stay that would not be currently acceptable (4.2 days).

The last limitation we will discuss is related to the choice of the model itself. As outlined earlier, the Markov technique is one method among a group of dynamic models that allow the modeling of events over time. This characteristic is essential for the problem at hand as relapse occurs gradually with time, and the costs and disutility of medical therapy also accrue with time. Given that time and the current clinical state are likely the most powerful predictors of outcome, and that previous events have little impact, we felt this model was appropriate. The main limitation of the model, lack of memory, makes it impossible to model the “snowballing” effect wherein a patient who, for example, receives a dilatation for gas-bloat symptoms is more likely than others to require a subsequent dilatation, or that a patient who requires a conversion to an open laparotomy is more or less likely to go on to require medications. This argument can be made in the opposite direction, since patients who have fared well in the first year after surgery, for example, may tend to be at lower risk for subsequent complications. However, since costs and utilities are bounded below but unbounded above, there remains the possibility of a bias against the medical arm. The model inherently makes the groups more homogeneous and thereby tends to overestimate precision by not taking into account these rare outliers. Regardless, the degree to which these previous events influence most future events is not accurately known, and so modeling memory for these type of events, even if it were possible, would not have borne much validity.

Mathematical simulations must be regarded as such and are in no way meant to replace life experience. In addition, one is reminded from the upper end of the 95% interpercentile ranges derived from Monte Carlo analysis that surgical therapy can be debilitating and expensive when serious complications occur. Although in our model the laparoscopic approach appears to be on average more cost effective than long-term medical therapy, this does not predict the best approach for the individual. From a societal perspective, however, our model’s conclusions are highly relevant. Clinical judgment, patient preference, and patient selection are all of obvious importance.

While it is possible to model, from a macroeconomic perspective, the choices available in the long-term management of GERD, patients and their physicians will make the ultimate decision. They will derive their preferences from considering the adverse effects, costs, and anticipated long-term outcomes. To date, little work has been done on patient and physician attitudes to surgical versus medical management of GERD. In our experience, most primary care physicians regard LNF as expensive, invasive, and excessive. ¹⁰⁰ Our work has clearly demonstrated that these attitudes should be re-examined. In Canada, antireflux surgery is increasing in frequency in regions where the majority of procedures are performed laparoscopically. ¹⁰¹ We maintain that LNF is a cost-effective option for the treatment of erosive reflux esophagitis. One must be cautioned that these results cannot and should not be extrapolated to patients with either endoscopy-negative GERD or mild (grade I) esophagitis. When situations require long-term PPI maintenance and patient-specific factors such as age, surgical risk, and individual preference are favorable, LNF should be considered. Final data from several randomized trials in progress comparing medical therapy and antireflux surgery in the laparoscopic era are awaited.

Acknowledgments

The authors thank Janet Kluchky from the Gray Nuns Community Hospital and Health Center for her help in compiling the costs of supplies, as well as operating room and recovery room costs, associated with LNF.