Introduction

Key to management of neurogenic lower urinary tract dysfunction is preservation of renal function by avoidance of elevated storage pressure. The first-line treatment for achieving low pressure urine storage and appropriate urine evacuation is combination oral anticholinergic treatment and intermittent catheterization. There are a substantial percentage of patients with depressed bladder compliance, for whom anticholinergic treatment is ineffective. These individuals often have upper tract dilation and recurrent urinary infections requiring recurrent hospitalization. Augmentation cystoplasty (AC) has historically been an effective method for establishing a high capacity, low pressure reservoir and securing upper tract drainage, with intermittent catheterization frequently necessary as an adjunct for achieving optimal urinary continence. While effective, surgery is invasive and carries high complication and reintervention rates [1].

Detrusor injections with botulinum toxin type A (BTX-A) were first applied in March 1998 by Stohrer for treatment of neurogenic detrusor overactivity (NDO) [2]. It has since reached rapid acceptance due to its simple administration, reliable efficacy and lack of adverse events. A single dose of 200–300 U of intra-detrusor BTX-A injection has been successfully used to treat neurogenic detrusor overactivity for up to 9 months in a number of patients in various countries [3]. Prior studies indicate that repeat injections do not cause damage to the detrusor muscle or decrease detrusor compliance over time [4, 5]. Yet, the price of the toxin and inherent repetitive administration of BTX-A injections may have formidable cost implications. Long-term management requires understanding of the cost differential between these two options. We conducted a cost analysis of treatment options for refractory NDO. We estimated the average initial treatment costs and cumulative 5-year costs of BTX-A injections and AC for antimuscarinic refractory patients.

Methods

The base example for this cost analysis model was an individual with a neurogenic bladder and associated detrusor overactivity for whom anti-cholinergics and intermittent catheterization were not sufficient at maintaining safe bladder pressures and continence. These individuals often have decreased bladder compliance (mechanical elasticity), elevated storage pressures (>40 cm H2O) and detrusor overactivity associated with upper tract dilation and recurrent urinary tract infections. The example case does not include patients with decreased compliance from non-neurogenic causes or neurogenic patients in whom urodynamic evaluation reveals no detrusor activity.

The primary analysis was conducted from the perspective of the health care payor. Indirect costs, such as time lost from work, pain, emotional and personal distress were not included in this model. To this end, the patient or societal perspective was not assessed using this model.

This model included two initial treatment strategies: AC and BTX-A injections. It was assumed that all patients had a neurological cause for incontinence and the same initial work-up (laboratories, urine culture, renal ultrasound, post-void residual, cystoscopy, videourodynamics). Since these costs were the same across both groups, no incremental effect was expected on total costs. Figure 1 contains a schematic diagram of the pathways for each of the treatment strategies. For both treatments, it was assumed that total initial costs include operating room and recovery room costs, cost associated with hospital stay, and normal post-operative follow-up care that were within the global service coverage.

Fig. 1
figure 1

Decision tree for cost analysis model. Asterisk inside the figure indicates the treatment algorithm continues along augment pathway

Full size image

Based on our practice, treatment with BTX-A injections assumed that 300 U of intradetrusor BTX-A injections using the technique described by Stohrer et al. [2] was given every 6 months over a 5-year period (total of 10 injections) when successful. In general, non-neurogenic DO patients were maintained at lower levels (200 U) of toxin administration, which could have cost implications. We did not explore the costs associated with other doses of toxin administration.

All dollar amounts were reported in 2008–2009 US dollars. Given the relatively short time frame and the fact that most of the costs were incurred in the first year of the model, discounting was not employed. Average reimbursement (physician and facility) from up to eight insurance carriers were used to estimate costs associated with initial BTX-A and AC. These costs include the laboratory work, radiologic evaluation, drugs, hospital stay, surgical care and follow-up provided during the appropriate global service coverage. Average costs of complications outside of the global service coverage were calculated based on our standard practices. The daily cost of antibiotics (fluoroquinolone), $12.63, anticholinergics (ditropan), $1.00, and anti-motility agents (immodium), $1.68, was obtained through drugstore.com. Radiologic costs (CT of abdomen and pelvis without or without contrast), $4,215, were based upon 2008–2009 reimbursements to Vanderbilt Hospital from local insurance carriers. Average hospital stay costs varied depending on the acuity of the care. Calculation of costs associated with UTI following BTX-A injection used an average of treatment for a single UTI (7 days) and daily antibiotics for prevention of chronic UTI over 5 years. Similarly, calculation of costs associated with AUR following BTX-A injection is an average of a transient (7 day) episode of urinary retention and daily clean intermittent catheterization. Following AC, the cost of UTI was an average of simple (daily antibiotic) and complex UTI costs (2 days regular room hospitalization + CT + 7 days of antibiotics). The cost of bowel dysfunction was calculated as an average of simple (daily anti-motility agent) and complex (3 days regular room hospitalization + CT). Complex UTI and complex bowel dysfunction both assume that the population had chronic symptoms requiring daily medications with an acute exacerbation requiring hospitalization. Costs associated with management of AC surgical complications (SBO, stone formation and perforation) were based on average reimbursements by all local insurance companies.

A decision-analysis model was developed using TreeAge Pro Healthcare 2009 Software, Inc, Williamston, MA. Successful treatment with either AC or BTX-A and no associated complications over 5 years was assigned a payoff value of one, with all other outcomes assigned a payoff of zero. The tree was then rolled back and the average cost for each treatment arm was determined. One-way sensitivity analyses were performed on all variables in the model. Selected two-way sensitivity analyses were performed based on the results of the one-way sensitivity analysis. This provided better understanding of the variables, their interaction and applicability in a clinical setting.

Results

Probabilities and ranges used for constructing the model are included in Table 1. Based on the author’s clinical experience and literature review, we assumed that 50% (0.2–0.7) of patients receiving 5 years of BTX-A will require management for at least one episode of urinary retention and/or urinary tract infection (UTI) [6, 7]. In this strategy, 10% [8] (0.01–0.15) of patients will fail BTX-A injections and either receive anticholinergics or undergo AC. A 40% (0.02–0.65) complication rate, broken down to 65% (0.4–0.75) medical complications (chronic UTI, bowel dysfunction) and 35% (0.15–0.4) surgical complications (small bowel obstruction, stone formation, bladder perforation), was used for calculating initial AC treatment costs. Among the group who has no initial complications, rates in the literature indicate that 1.4% (0.00–0.02) will go on to have reaugmentation [9].

Table 1 Estimates of probabilities used in model and ranges tested in sensitivity analysis
Full size table

Cost estimates and ranges used for sensitivity analyses are presented in Table 2. The average physician and facility reimbursement for one round of BTX-A is $461.56 and $2,485.27, respectively. The average physician and facility reimbursement for AC is $2,946.35 and $22,095.18, respectively. Roll-back analysis determined that BTX-A treatment was the less expensive option over 5 years, costing $28,065. This was in comparison with AC which cost $33,272 and assumes that success for both groups represented the same clinical state. One-way sensitivity analyses were performed on all variables in the model over clinically appropriate ranges from the perspective of the facility and surgeon. The values that were used are presented in Tables 1 and 2. The model appears robust, as it is only sensitive to change within a reasonable clinical range for one variable: months of repeat BTX-A. Our roll-back analysis assumed the durability of BTX-A effect to be 6 months on average (10 shots of botox are given over 5 years). The threshold value for the durability of BTX-A injections was found to be 5.1 months. If the durability of botox is greater than 5.1 months, then BTX-A is the preferred treatment option. If the durability is less than 5.1 months on average, AC becomes the favored treatment modality. This is presented in Fig. 2. Another variable that is sensitive to change was complication rate after augmentation. Our model was based on an AC complication rate of 40%. One-way sensitivity analysis revealed a threshold value of 14%. If the actual probability of complication rate drops below 14%, including both medical and surgical complications, AC is favored. This is presented in Fig. 3. The model is additionally sensitive to facility and surgeon costs of BTX-A and facility cost of AC, with threshold values of $3,027.74, $1,004.03 and $17,100, respectively.

Table 2 Estimates of costs used in model and ranges tested in sensitivity analysis
Full size table
Fig. 2
figure 2

One-way sensitivity analysis of the durability of the effectiveness of botox (in months)

Full size image
Fig. 3
figure 3

One-way sensitivity analysis of the probability of a complication after augmentation cystoplasty 

Full size image

Discussion

BTX-A and AC are the two treatment options for antimuscarinic-resistant NDO. Only two studies have analyzed the cost-effectiveness of BTX-A injections based on clinical outcomes. A study of 63 cases of neurogenic DO found BTX-A a more cost-effective option than standard of care in a UK secondary care setting for patients not planned for surgery in the time frame of the analysis [6]. A German group found BTX-A injections decrease the costs of incontinence aids and UTI treatment by 50% [7]. No study to date has compared cost implications of BTX-A to gold standard management with augmentation cystoplasty. This study shows that BTX-A is a less expensive option than AC over a 5-year period when success is defined as the same clinical state.

Following the FDA approval of BTX-A in 2,000, questions arose regarding the safety, durability and changes to bladder compliance with repeated injections. In the largest retrospective, multicenter European clinical trial [3], 200 neurogenic bladder patients were reported to have an increase greater than 50% of mean volume to first reflex detrusor contraction and maximum cystometric capacity, with proportional decrease in maximum detrusor pressure. This effect exceeded 6 months. Similar results with longer duration of effectiveness (12 months) were reported by Del Popolo et al. when 500–1,000 U were given [8]. The clinical acceptance of BTX-A has increased with studies confirming the efficacy and safety of the toxin [4, 9–13]. A systematic review of 18 articles [14] evaluating the efficacy and safety of BTX-A in patients with neurogenic detrusor overactivity found injections effective for a minimum of 12–39 weeks [3, 4, 15–18] with significant quality of life improvements up to 12–24 weeks [15, 19–21]. Those studies finding faster decline in efficacy [3, 22] (<24 weeks) [3, 17, 23] may apply to patients with lower baseline bladder compliance [22]. Our study found AC to be the cheaper option if the durability of BTX-A is less than 5.1 months. Preoperative urodynamics may help in differentiating those patients who would benefit from AC over BTX-A from a cost perspective. BTX-A may become less effective in patients with progressively decreasing bladder compliance secondary to changes in mechanical elasticity, rather than from the dynamic nature of overactivity. This may have cost repercussions beyond the scope of this analysis. Clearly, the timing and magnitude of reinjections and rationale (symptoms, urodynamic changes) for reinjection are essential in providing an effective yet cost-conscious solution. In longer-term follow-up (10 years), Pannek et al. [24] found evidence of impaired detrusor contraction strength over time and 25% rate of requiring surgical intervention. Our sensitivity analysis ranging from 5 to 60% for probability of botox failure did not find this variable to be significant in the cost differential. This indicates that even with injection-related changes in detrusor function, unless reinjections are necessary more than every 5.1 months, BTX-A remains cheaper than AC. Two-way sensitivity analysis of months until injection and BTX-A facility cost, not surprisingly, finds increasing cost of BTX-A (greater than $2,900) causes AC to be favored.

Benefits of AC are well established, as are the associated high complication and reintervention rates [1, 25–27]. This model was based on a 40% overall AC complication rate from the series with the longest mean follow-up [1]. There is a great deal of variability in these rates within the reported literature, likely due to a lack of standardization. Sensitivity analysis reveals that AC is only cheaper than botox if the complication rate (including medical and surgical) is less than 14%. Even when performed at high volume academic centers, it is unlikely that the complication rate will be this low due to the population involved and nature of surgical intervention. The model in this study is based on the standard of care within our institution and from the reported literature. While costs were calculated based on facility and physician-related dollars, study was limited by inability to define costs to the individual or to the society. Specifically, we do not include indirect costs, such as time lost from work, lost wages, effects on personal relationships, costs of incontinence products, or quality of life impact. Calculations of medical complications were based on an average of extreme conditions (i.e. isolated UTI and complex UTI in setting of prophylaxis), which artificially simplifies the range of presentations. This study is also limited by the data available in the literature. There is a lack of standardization in the reporting of complication rates following AC. There are no large, randomized, prospective trial to evaluate efficacy, safety and duration of BTX-A effects. There remains variability in dosage, number and location of BTX-A injection sites. Our model was based on a 5-year perspective. BTX-A data are not as mature as AC data and the long-term duration of effects are still being accumulated. That being said, it is possible that the results of the analysis would have been quite different over a longer time horizon.

Conclusions

Bladder dysfunction associated with neurogenic detrusor overactivity impairs quality of life and often poses a threat to the upper urinary tracts, especially when the condition is unaddressed. Following failure of conservative measures, AC or BTX-A are used when some detrusor function remains. Due to the high complication and reoperation rate of AC, BTX-A injections are increasingly being used for management. This is the first study to date to perform a cost analysis of these two options. BTX-A proved to be cheaper at durations greater than or equal to 5.1 months and AC was cheaper when the cost of BTX-A increased or the AC complication rate was dropped below 14% in the model. Due to the variability in practices and detrusor compliance, there will be a subset of patients who may need BTX-A more often due to worsening of the detrusor function. Urodynamic parameters may help with counseling and the decision to pursue a surgical intervention to avoid excessive reinjections and greater financial burden. This information is a valuable adjunct in the clinical decision tree of NDO.