Unplanned Health Care Utilization after Ureteroscopy with Ureteral Stenting: Results from the Study to Enhance Understanding of Stent-Associated Symptoms Cohort.

STENTS is a multicenter, prospective, observational cohort study of individuals undergoing ureteroscopy and ureteral stent placement after treatment of a ureteral or renal stone. A description of the STENTS protocol has previously been published. 12 Participants aged 12 years and older with a planned ureteroscopy for stone treatment were recruited from four clinical centers. All participants were prospectively enrolled after institutional review board approval and informed consent. Participants aged 17 years or younger provided their informed assent, and their parents provided parental permission. Exclusion criteria were an indwelling ureteral stent within the preceding 60 days, concomitant shockwave lithotripsy or percutaneous nephrolithotomy, conditions resulting in neurogenic bladder dysfunction, anatomic urological abnormality resulting in abnormal bladder sensation, and renal transplantation. Vulnerable populations (e.g., prisoners or individuals with cognitive impairment that would impact their ability to participate in the protocol) were also excluded. Participants completed baseline questionnaires prior to surgery that recorded individual characteristics, medical and stone history, existing chronic pain conditions, and medication use. Participants completed multiple outcome instruments assessing pain intensity, pain interference, and urinary symptoms before surgery (baseline), on postoperative days 1, 3, and 5, on the day of stent removal, and 30 days after stent removal. Questionnaires were self-administered and completed via electronic format, or paper copies if preferred. Trained research coordinators recorded intraoperative data at the time of ureteroscopy, including stone features, details of ureteral instrumentation, irrigation type, and stent characteristics, which were verified and confirmed by the treating urologist. At time of stent removal and 30 days after stent removal, participants were queried as to whether they participated in any interaction with a healthcare provider since their surgery. All healthcare utilization events within 30 days of surgery were evaluated. A single reviewer (BJ) reviewed the medical record for each encounter to determine the nature of the encounter as whether it was related surgery and/or stent. Any event that could be classified as a Clavien-Dindo complication was automatically associated with the surgery. Unplanned encounters related to the surgery (UEs) were isolated from expected clinical encounter (e.g. stent removal visit) or unrelated encounter (e.g. previously scheduled primary care visit or encounter for medication refill). Data were analyzed to determine which factors were associated with a higher risk of having a UE related to surgery. Participants’ characteristics were summarized as means with standard deviations for continuous variables, and as counts with percentages for categorical variables. Comparisons between groups were made using two-sample t-test for continuous variables and Pearson’s chi-square test or Fisher’s exact test for categorical variables. To identify potential variables linked to a UE, we prespecified 20 candidate variables based on existing literature, expert consensus, and clinical relevance. These variables included demographic information, medical history, and characteristics of the procedures and stents utilized. Univariable logistic regression models were used in estimating the odds ratios and 95% confidence intervals to evaluate the strength of associations between the prespecified variables ( Supplemental Table 1 ) and the primary outcome of UE. When the standard logistic regression model failed due to a small number of UE events for certain covariates, a penalized approach using Firth’s logistic regression method was utilized through the R package “logistf.” To further explore the associations of these candidate variables with UE, an initial full multivariable logistic regression model was developed, incorporating all 20 candidate variables. The final model was selected using a stepwise variable selection algorithm based on the Akaike Information Criterion. Missing data were not imputed. All analyses were conducted using R software, version 4.4.1. 13

A total of 484 participants (451 adults) underwent ureteroscopy for USD (424 unilateral, 60 bilateral). Mean age was 49 ± 17 years, and 47% were female ( Table 1 ). All postoperative clinical encounters were analyzed to determine whether they were UEs related to surgery. Within the cohort, 49 participants (10%) had at least one UE deemed related to the urological procedure within 30 days of surgery (median encounter 5 days after surgery, interquartile range: 2–16). The UE was an emergency room visit with or without admission in 24/49 (49% of UEs) of participants compared to an outpatient visit for the remaining 25/49 (51%). Age, race, sex, or previous stone history were not statistically different for those with or without a UE ( Table 1 ). Thirty-nine percent of the cohort had at least one chronic pain condition ( Table 2 ). On univariable analysis, the presence of a chronic pain condition, intraoperative ureteral injury, and ureteral dilation were all associated with a higher incidence of a UE, while use of manual pressure irrigation was protective ( Supplemental Table 1 ). Operative characteristics including variables such as surgery duration, bilateral procedures, use of a ureteral access sheath, or stent diameter did not differ between the groups. For those with a self-reported pain condition, a UE occurred 17% of the time compared to 6% for those without a pain condition (p < 0.001). For participants with a noted ureteral injury, a UE occurred in 24% compared to 10% for those without an injury. In this cohort, 29 ureteral injuries were noted (24 grade I, 4 grade II, and 1 grade III). To determine which covariates were independent predictors of a UE, a multivariable logistic regression model based on stepwise selection procedure was performed (443 of 484 included in the final multivariable analysis, due to missingness). This model demonstrated presence of a chronic pain condition to be the only variable to be independently associated with a UE ( Table 3 ). The type of ureteral stent, ureteral injury, and gravity irrigation were not found to be associated with a UE.

The prevalence of unplanned healthcare utilization related to ureteroscopy and stenting for USD in this large, multicenter cohort was 10%. These encounters are costly and burdensome to the patients. Understanding these UEs as well as the patient characteristics that increase their risk represent an unmet need in the field of surgical stone managements. Those with a chronic pain condition were almost three times more likely to require a UE. Additional research is needed to determine how to mitigate the risk in these patients. Potentially more thorough counseling or a more intense clinical care pathway would reduce the risk of UEs. Ultimately, identifying patients at highest risk is the first step in reducing UEs and enabling proactive targeted interventions.

Despite significant advances in technology for the surgical treatment of USD, ureteral stenting remains a significant source of morbidity and unplanned healthcare utilization events. 8 We analyzed potential predisposing factors associated with unplanned utilization of the healthcare system using a prospective large cohort of well-characterized individuals undergoing ureteroscopy with stone treatment and stent placement. The comprehensive prospective data collection allowed for robust analysis of the population with a very common urological condition. Importantly, we noted that 10% of our cohort experienced a UE within 30 days following ureteroscopy and identified the presence of a chronic pain condition and occurrence of a ureteral injury as independent predictors of UE. Our finding of a 10% UE rate is consistent with the previously published Marketscan study by Scales and colleagues. 4 A more recent prospective cohort study of post-ureteroscopy emergency room visits demonstrated a similar rate of 8% compared to our rate of 5% for emergency visits (excluding outpatient visits). 14 Following implementation of an early recovery after surgery protocol, the rate of these emergency room visits remained the same. These events place a strain on the healthcare system and take a mental, physical, and financial toll on the patients. With a better understanding of the cause and contributing factors of stent pain–related UEs, efforts to mitigate them can be better directed. For the well-selected patient, more thorough patient counseling regarding expected symptoms or additional pharmacological therapy may be beneficial. In a prior STENTS analysis, we found that patient factors (i.e., younger age, depression/anxiety, chronic pain condition, severe pain with a prior stent) were the most strongly associated with increased stent-associated pain. 11 The primary drivers of stent discomfort are quite elusive. It is difficult to uncouple the symptomatology of the stent itself from that of the endoscopic procedures (e.g., ureteral access sheath, intrarenal pressure increase, temperature, stone extraction). Indeed, a recent case control study of ureteral stent for urolithiasis compared to renal transplantation demonstrated significantly worse pain, urinary symptoms, and opioid use in the USD group. 15 Although a transplanted ureter does not have the same innervation as a native kidney, the bladder-related symptoms were less bothersome. These findings suggest that the “stent” symptoms may be due, in some part, to the surgery/ureteroscopy itself. However, a potential limitation of this study is that pretransplant end stage renal transplantation patients may be oliguric with altered bladder function. Similarly, a foley catheter following the transplant surgery may mitigate some of the stent symptoms. A recent prospective randomized study suggested that stent removal at day 3 reported better urinary and pain scores without any difference in unplanned encounters. 16 Similarly, our data indicates that length of stent dwell time similarly was not associated with rate of UEs. Several operative parameters were evaluated to determine whether differences in the endoscopic portion of the procedure were associated with incident UEs. Neither the use and size of a ureteral access sheath, bilaterality of the procedure, need for ureteral dilation, type of irrigation device, length of procedure, nor size of stone portended a high risk of UEs. Although presence of a ureteral injury was significant on initial univariable analysis, it did not demonstrate an independent association with rate of UEs in the multivariable model. Determination of ureteral injury in this cohort was based on visual inspection of the ureter upon retraction of the sheath/scope. Reported rates in the literature have a wide range from 6% 16 to 72%, 18 underscoring the subjective nature of the grading system. The identification of a ureteral injury may influence the surgeon’s decision regarding stent dwell time or need for a staged procedure. Stent characteristics such as its length and diameter, presence of a complete proximal or distal coil, amount of stent in the bladder, or use of a string on the stent were not associated with UEs. A recent systematic review of the use of a stent tether indicated that the use of stent tethers decreased the length of time the stent was in placed with no difference in complications, urinary tract infections, or urinary symptoms. 19 Previous studies suggest that a longer stent with the proximal end in the upper calyx and the distal end crossing the midline of the bladder generated worse urinary urgency and dysuria and lower quality-of-life scores with no difference in flank pain. 20 However, a recent study of 73 patients found no significant influence of intravesical stent position on associated morbidity, nor did it find a significant association between urinary symptom scores. 21 A systematic review concluded general consensus on the lack of association between stent length and stent-related flank pain. 7 To our knowledge, this is the first study to evaluate associations between stent characteristics and unplanned clinical encounters. Interestingly, initial analysis demonstrated an increased risk specifically for the Cook Black Silicone Stent, which is associated with lower postoperative stent discomfort. 22 However, after the multivariable stepwise regression, stent type was not associated with a UE. It has previously been suggested that a silicone stent is preferred for patients with known pain conditions or previous issues with stents as a pain mitigation technique. 23 Therefore, surgeons in this cohort may have been more inclined to use a silicone stent for patients with a history of stent intolerance. By far the strongest association with a UE in this cohort was the presence of a chronic pain condition. These pain conditions were largely non-urologic in nature. In the entire cohort, 39% of participants identified at least one pain condition, with arthritis, joint pain, and back pain being the most prevalent. Prior analyses of the STENTS cohort demonstrate persistence of associations between chronic pain conditions and greater stent-associated symptoms, even after controlling for medication use. 24 Chronic pain conditions are known to be associated with central sensitivity that, in turn, worsens the postoperative course following stone treatment. Lai and colleagues demonstrated that participants in this STENTS cohort who were more sensitive to a fixed stimulus by quantitative sensory testing (QST) had higher postoperative pain intensity. Moreover, those with chronic pain conditions were more likely to have preoperative hypersensitivity on QST. 25 Therefore, there may be a component of central pain sensitivity that predisposes one to UEs following surgery. Additionally, there may be a greater familiarity with the health system that lowers the barrier to care-seeking among patients with chronic pain conditions. This study has several strengths. The STENTS cohort is robust and comprehensive, spanning four clinical centers across the United States. The cohort is well described, and follow-up data have little missingness. The cohort is well characterized with respect to psychosocial characteristics, baseline clinical data, granular procedural details, outcomes, and symptom assessment using multiple validated instruments. Additionally, the sample size is considerable. However, several weaknesses also exist. Postoperative clinical encounters were not a primary endpoint of the STENTS study. As thorough as the data collection was, there is still some missingness that required censorship in the analysis. While the rate of missing data is low, we cannot entirely exclude the possibility of an impact on the results. Several of the covariates in the study are correlated, which can introduce error in the analysis; however, our statistical methods were designed to mitigate this error. The cohort was heavily engaged in the study and therefore may not reflect a standard ureteroscopy patient with less postoperative contact. Finally, all sites were tertiary academic stone referral centers, and the care received may differ in unmeasured ways from that in the community.

The placement of a ureteral stent after routine ureteroscopy for kidney stone treatment is a common occurrence, 1 although it is considered optional by several guidelines 2 — surgeons elect for stent placement in as many as 80% of routine cases. 3 However, patients often do not tolerate stents well due to a myriad of stent-associated symptoms (pain, hematuria, dysuria). 4 , 5 , 6 While the exact mechanism of stent-related discomfort is unclear, proposed explanations include mucosal irritation from the distal stent coil, increased bladder spasms, ureteral aperistalsis, and reflux of urine. 7 When compared to nonstented individuals following routine ureteroscopy, patients with stents have demonstrated significantly higher rates of pain and urinary symptoms. 8 Patients experiencing these stent symptoms or complications from stone surgery often present to the emergency room or request an urgent outpatient clinic visit with their provider. While in some instances, these unplanned encounters (UEs) are needed to diagnose and treat a complication, some encounters may simply be due to poor tolerance of the stent. These events place a strain on the healthcare system as well as take a mental, physical, and financial toll on the patients. To that end, many groups have focused on stent-specific characteristics such as length, diameter, and location of placement to help explain poor stent tolerability. 7 The degree to which patient symptoms lead to increased use of healthcare resources is not clear. For example, evidence to suggest that stent placement directly increases emergency department visits and hospital admissions is conflicting. 9 , 10 Recent research from USDRN has demonstrated that the main risk factors for stent-associated symptoms are primarily patient-related. 11 To gain further insight into the potential factors related to stent-associated symptoms, the National Institute of Diabetes and Digestive and Kidney Diseases Urinary Stone Disease Research Network conducted the STudy to Enhance uNderstanding of sTent-associated Symptoms (STENTS), a prospective observational cohort study of adolescents and adults undergoing ureteroscopy with ureteral stent placement to treat urinary stone disease (USD). 12 Analysis of the data collected during STENTS provides a unique opportunity to analyze the impact that ureteral stent placement following ureteroscopy has on healthcare utilization within 30 days of surgery.