Objective
The purpose of this study was to determine the cost-effectiveness of serial stenting vs ureteroscopy for treatment of urolithiasis during pregnancy as a function of gestational age (GA) at diagnosis.
Study Design
We built decision analytic models for a hypothetical cohort of pregnant women who had received a diagnosis of symptomatic ureteral calculi and compared serial stenting to ureteroscopy. We assumed ureteral stent replacement every 4 weeks during pregnancy, intravenous sedation for stent placement, and spinal anesthetic for ureteroscopy. Outcomes were derived from the literature and included stent infection, migration, spontaneous kidney stone passage, ureteral injury, failed ureteroscopy, postoperative urinary tract infection, sepsis, and anesthetic complications. Four separate analyses were run based on the GA at diagnosis of urolithiasis. Using direct costs and quality-adjusted life years, we reported the incremental costs and effectiveness of each strategy based on GA at kidney stone diagnosis and calculated the net monetary benefit. We performed 1-way and Monte-Carlo sensitivity analyses to assess the strength of the model.
Results
Ureteroscopy was less costly and more effective for urolithiasis, irrespective of GA at diagnosis. The incremental cost of ureteroscopy increased from –$74,469 to –$7631, and the incremental effectiveness decreased from 0.49 to 0.05 quality-adjusted life years for a kidney stone diagnosed at 12 and 36 weeks of gestation, respectively. The net monetary benefit of ureteroscopy progressively decreased for kidney stones that were diagnosed later in pregnancy. The model was robust to all variables.
Conclusion
Ureteroscopy is less costly and more effective relative to serial stenting for urolithiasis, regardless of the GA at diagnosis. Ureteroscopy is most beneficial for women who received the diagnosis early during pregnancy.
Genitourinary complications are the leading cause of nonobstetric hospital admissions during pregnancy, with kidney stones affecting an estimated 1 in 244 to 1 in 2000 pregnancies. Many anatomic and physiologic changes that accompany pregnancy are associated with an increased risk of kidney stone formation. These include mechanical obstruction of the ureters, progesterone-induced reduction in ureteral peristalsis, and increased filtration rates of kidney stone promoters such as sodium, calcium, and uric acid. In addition, pregnant patients may be at increased risk of kidney stone–related complications, such as pyelonephritis or obstetric complications that include preterm labor and preterm delivery.
As in the general population, analgesia and monitoring for spontaneous passage are often the most appropriate initial treatments for acute renal colic in the pregnant patient. However, recent studies have found that as low as 47% of pregnant patients with confirmed urolithiasis are able to pass kidney stones spontaneously. If conservative treatment is unsuccessful or if there are absolute indications for intervention (eg, intractable pain, vomiting, fever), active treatment is warranted. Traditionally, antepartum ureteral stent insertion has been the most commonly used method to alleviate urinary tract obstruction temporarily in pregnant women, which allows pregnancy to continue to term without more invasive treatment methods such as ureteroscopy. Because of changes in urine composition that occur during pregnancy, there is an increased rate of stent encrustation, which necessitates stent replacements every 4-6 weeks. Antepartum stent placement has 2 advantages: it allows for immediate relief of the obstruction and subsequently leads to passive dilation of the ureter, potentially facilitating successful ureteroscopy after delivery. However, stenting is not entirely benign. Indwelling ureteral stents can become infected, migrate, or cause significant pain. The pain that is associated with indwelling stents is often so severe that many patients report higher total pain scores than those with acute renal colic; 80% of patients with a stent report a significantly reduced quality of life.
In contrast, immediate ureteroscopy, as is performed in the nonpregnant patient, eliminates the need for antepartum serial ureteral stent placement and provides a definitive surgical treatment. However, there remains controversy regarding the use of ureteroscopy during pregnancy. Ureteroscopy requires a higher level of anesthesia and increased procedure time relative to ureteral stent placement and carries a risk of ureteral perforation or injury. Additionally, without preoperative stenting, antepartum ureteroscopy may be less successful relative to postpartum ureteroscopy after serial stenting. In response to these concerns, several studies have demonstrated that ureteroscopy is highly effective in pregnancy and can be performed under spinal anesthesia with minimal reported risks to the mother and fetus. Multiple small studies have also shown no significant differences in complications after ureteroscopy in pregnant women, compared with nonpregnant women. Additionally, the American Urological Association treatment guidelines now recognize ureteroscopy as a treatment option to be considered for pregnant patients with urolithiasis.
It remains unclear whether antepartum serial stenting or ureteroscopy is superior for the treatment of kidney stones during pregnancy and how the gestational age (GA) at diagnosis affects the risks and benefits of each treatment modality. Because ureteroscopy is more successful after stent placement, we hypothesized that stenting later in pregnancy may be the preferred approach, although ureteroscopy earlier in pregnancy may prove to be more beneficial. The objective of this study was to determine the cost-effectiveness of serial stenting vs ureteroscopy for the treatment of urolithiasis in pregnant women as a function of GA at diagnosis.
Materials and Methods
To compare patient outcomes that result from ureteral stent placement vs ureteroscopy, we developed a decision analytic model for a hypothetical cohort of women with a symptomatic ureteral stone during pregnancy ( Figure 1 ). We conducted the analysis from a payer’s perspective and measured effectiveness as quality adjusted life years (QALYs.) QALYs were calculated with the use of life expectancy and utility values, which are defined as numeric judgments of the desirability of specific outcomes. We calculated incremental costs, incremental effectiveness, incremental cost-effectiveness ratio, and net monetary benefit (NMB). Incremental costs are defined as the difference in the mean cost between the 2 strategies. Similarly, incremental effectiveness is the difference in the mean effectiveness (ie, QALYs) between the 2 strategies. We calculated the incremental cost-effectiveness ratio as the incremental cost divided by the incremental effectiveness and compared this value to a willingness to pay (WTP) threshold of $50,000 per QALY to determine whether ureteroscopy was cost-effective compared with ureteral stenting. Although there is no consensus on a WTP threshold, $50,000 was chosen because it is frequently used in cost-effectiveness health care analyses that originate in the United States. We also assessed the cost-effectiveness of ureteroscopy with the use of NMB, which is defined as the WTP × incremental effectiveness – incremental cost. If the NMB of ureteroscopy was greater than zero, ureteroscopy was cost-effective compared with stent placement.
To develop the model, we used TreeAge Pro software (TreeAge Software, Inc, Williamstown, MA). We assumed that patients who underwent stent placement received intravenous sedation, that stents were changed every 4 weeks during pregnancy, and that there were no changes in symptoms over the duration of the stent. In addition, we assumed that, if stent migration occurred, the stent was replaced and that, if ureteral injury occurred during ureteroscopy, serial stents were placed every 4 weeks for the remainder of the pregnancy. If the ureteroscopy procedure failed (ie, kidney stone recurrence or persistence), we assumed that a stent would be placed for 2 weeks, followed by a second attempt at the ureteroscopy procedure. All pregnancies were assumed to result in delivery at 40 weeks of gestation. We also assumed patients who underwent ureteroscopy received spinal anesthesia. We based this assumption on findings that supported the safety and efficacy of spinal anesthesia for ureteroscopy. Additionally, regional anesthesia during pregnancy remains the preferred alternative because the implications for general anesthesia for long-term neurodevelopment of the fetus remain unclear.
We set the GA of kidney stone diagnosis at 4 different time points: 12, 20, 32, and 36 weeks of gestation. We conducted separate analyses for each diagnostic time point. To calculate stent costs and disutility at the varying time points of diagnosis, we multiplied the cost and disutility of stent placement by the number of stents that were needed to reach delivery, assuming stent replacement every 4 weeks.
We obtained all modeling probabilities from the literature ( Table 1 ). If multiple estimates were available in the literature, we calculated a mean estimate by weighting the individual estimates based on sample size. We included this weighted average as the parameter value in the base case analysis.
| Variable | Baseline, % | Range, % | References |
|---|---|---|---|
| Stent complications | |||
| Infection | 9.86 | 6.00–12.30 | |
| Migration | 2.38 | 1.20–10.00 | |
| Spontaneous kidney stone passage | 10.00 | 2.10–26.50 | |
| Ureteroscopy complications a | |||
| Postprocedure temporary stent | 56.70 | 42.90–72.90 | |
| Failure of ureteroscopy without preoperative stent | 3.04 | 0.00–15.35 | |
| Failure of ureteroscopy with preoperative stent | 0.00 | 0.00–17.60 | |
| Urinary tract infection | 1.14 | 0.60–5.20 | |
| Ureteral injury b | 1.69 | 0.86–6.70 | |
| Sepsis | 0.51 | 0.30–2.30 | |
| Ureteroscopy anesthesia complications c | |||
| Postdural headache | 2.20 | 1.30–9.60 | |
| Aseptic meningitis | 0.17 | 0.00–0.90 |
a A lower ureteroscopy failure rate was used for patients who underwent postpartum ureteroscopy after antepartum serial stenting
b Included perforation, avulsion, and stricture
c Complications from stent anesthesia were not included because of the relatively low occurrence and transient nature of intravenous anesthesia complications.
For stent placement outcomes, we included infection, stent migration, and spontaneous kidney stone passage. We based stent complications on elective stent placement and defined infection by the presence of fever and bacteruria. Because stent placement is primarily a temporizing procedure, we included the cost of postpartum treatment within the model for patients who underwent antepartum serial stenting. We assumed that all stented patients underwent an abdominal computed tomography scan to assess for persistent kidney stone after delivery. If spontaneous kidney stone passage occurred after serial stent placement during pregnancy, the patient underwent no further treatment. If no spontaneous kidney stone passage occurred, the patient underwent ureteroscopy 2 weeks after delivery. A higher ureteroscopy stone-free rate was included for postpartum ureteroscopy, relative to antepartum ureteroscopy based on the documented improved success of ureteroscopy after preoperative stenting.
For ureteroscopy outcomes, we included temporary postoperative stent placement (for 1 week duration after ureteroscopy), urinary tract infection, sepsis, ureteroscopy failure without preoperative stent placement, ureteroscopy failure with preoperative stent placement, and ureteral injury. When available, we included complication rates that were specific to pregnant patients. However, we also incorporated studies that included nonpregnant patients because of the equivalent complication rates found between pregnant and nonpregnant patients after ureteroscopy. With regard to anesthesia complications, we included postdural puncture headache and aseptic meningitis after spinal anesthesia. We did not include complications of intravenous anesthesia because these complications tend to be rare and transient, likely having an insignificant effect on the outcome of the model.
The model included hospital, urologist, and anesthesiologist costs and the cost of intraoperative fetal monitoring with fetal nonstress test during ureteroscopy or stent placement ( Table 2 ). We based urologist costs on the Current Procedural Terminology from the American Medical Association. We calculated anesthesiologist costs using the Centers for Medicaid and Medicare Services anesthesia charge formula and derived costs for urinary tract infection, stent infection, computed tomography scan, and postdural headache from the literature. More specifically, we calculated postdural headache costs using the mean cost per patient and including the average costs and use of intravenous caffeine, blood patch, and hospital stay. Infection costs included both the cost of treatment and laboratory diagnosis. We obtained costs for inpatient procedures and diagnoses (ie, sepsis, meningitis) from the 2012 Hospital Cost and Utilization Project, a nationwide inpatient database that reports the national average costs for specific diagnoses and procedures. For outpatient procedures (ureteroscopy and stent placement), we used cost data from the NorthShore University HealthSystem financial department. To convert all costs to 2014 dollars, we used the medical care component of the Consumer Price Index. Additionally, we discounted all future costs and utility values at a 3% annual rate.
| Variable | Baseline, $ | Range, $ | References |
|---|---|---|---|
| Hospital costs | |||
| Stent | 10,422 | 5211–15,633 | — b |
| Ureteroscopy | 13,403 | 6702–20,105 | — b |
| Computed tomography scan | 383 | 192–575 | |
| Fetal nonstress test | 624 | 312–936 | — b |
| Provider | |||
| Urologist | |||
| Stent | 177 | 89–266 | |
| Ureteroscopy | 474 | 237–711 | |
| Anesthesiologist c | |||
| Stent | 77 | 39–116 | |
| Ureteroscopy | 207 | 103–310 | |
| Outcome | |||
| Stent infection d | 1,573 | 786–2359 | |
| Urinary tract infection | 1,573 | 786–2359 | |
| Ureteral injury e | 13,621 | 6810–20,431 | |
| Sepsis | 14,401 | 7200–21,601 | |
| Postdural headache | 1,712 | 856–2568 | |
| Aseptic meningitis | 14,490 | 7245–21,735 |
a All costs were converted to 2013 dollars with the use of the Consumer Price Index
b Based on data from the NorthShore University HealthSystem financial department
c For anesthesiologist costs, the Medicare formula of “Charge = (base units + time units + modifier units) × conversion factor” was used; for ureteroscopy, the charge calculation was done with the use of 2 base units, 4 time units, and 2 modifier units; for ureter stent placement, the charge calculation was done with the use of 2 base units, 1 time unit, and 0 modifier units
d The cost of stent infection was estimated with the use of the cost of urinary tract infection
We obtained all utility values from the literature ( Table 3 ). To calculate QALYs, we combined the utility values, tolls, and the estimated life expectancy. We defined a full life expectancy as 56.8 years in the model based on the life expectancy of a 25-year-old woman (average age of first pregnancy). If a patient had multiple complications, we used the complication with the lowest utility value at any given time point to calculate the corresponding QALYs.
| Variable | Utility | Range | Time interval, d | References |
|---|---|---|---|---|
| Procedure | ||||
| Stent | 0.76 | 0.64–0.88 | Varied a | |
| Ureteroscopy | 0.90 | 0.85–0.95 | 7.9 | |
| Outcome | ||||
| Stent infection b | 0.64 | 0.46–0.82 | 7 c | |
| Urinary tract infection b | 0.64 | 0.46–0.82 | 7 c | |
| Ureteral injury | 0.60 | 0.40–0.80 | 30 | |
| Sepsis b | 0.64 | 0.46–0.82 | 7 c | |
| Postdural headache | 0.77 | 0.81–0.94 | 6 | |
| Aseptic meningitis | 0.65 | 0.48–0.83 | 8.5 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
