1 INTRODUCTION
Renal stone incidence and prevalence have been increasing in the United States and other parts of the world[
1]. Percutaneous nephrolithotomy (PCNL) is the standard management for complex renal stone disease[
2]. However, the procedure is elective, and patients consequently may experience significant delays in time to operation. In addition, the SARS‐CoV‐2 (COVID‐19) pandemic has led to delays in elective urologic surgical interventions[
3].
Although some studies have aimed to evaluate the impact of delaying kidney stone surgery on outcomes[
4,
5], few have examined the effect of delaying PCNL surgeries on hospital cost. Population‐wide research demonstrates that delays in PCNL surgery lead to increased emergency department visits, medical resource usage, and preoperative morbidity[
5,
6]. Factors associated with a longer time to stone surgery include being underinsured and minority status. The underinsured patients are more likely to revisit an emergency department (ED) three or more times, undergo two or more computed tomography (CT) imaging studies, and receive upper urinary tract decompression[
5]. However, few studies have evaluated delays in stone treatment among a medically underserved population, and, to our knowledge, none have reported factors associated with delays following the COVID‐19 pandemic.
In this study, we sought to investigate the effect that prolonged time to PCNL has on associated medical costs and healthcare resource utilization. We hypothesized that increased time prior to intervention in a county hospital system would be associated with higher hospital charges, more clinic and hospital visits, and more procedural interventions.
2 METHODS
2.1 Study design
A retrospective chart review of patients undergoing PCNL from April 2019 to December 2022 was performed using our institutional database at a single center in Houston, Texas. A total of 146 patients underwent PCNL at the institution during this period. Inclusion criteria were patients who underwent ambulatory urology encounters at the outpatient center and subsequently were scheduled for PCNL surgery. Exclusion criteria were patients with bilateral nephrolithiasis with surgery indicated for side contralateral to PCNL as well as patients <18 years old. Demographic data included age, gender, ethnicity, BMI, and insurance status (Table 1).
2.2 Data collection
Clinical data of interest included time from date of diagnosis to PCNL surgery (our exposure of interest); number of emergency department (ED) visits, hospital admissions, and clinic visits until surgery; number of CT scans, abdominal X‐rays, and renal ultrasounds until surgery; stent versus nephrostomy decompression, number of stent and nephrostomy‐related procedural interventions, presence of stent encrustation, and number of nephrostomy dislodgments documented.
Diagnosis date was determined by the day of the CT scan used to diagnose the kidney stone that required PCNL surgery. We included CT scans from the emergency room, inpatient, outpatient, and outside facility settings when possible. Groups were organized according to diagnosis‐to‐OR period in approximate tertiles rounded to the nearest month (less than 4 months, 4–6 months, greater than 6 months).
Our primary outcomes of interest were hospital charges, number of CT scans, and number of procedural interventions prior to surgery. Hospital charges were based on a dollar amount estimate obtained from data specific to the hospital in our study (Supporting Information: Table S1 and Figure S1). Due to difficulty in obtaining all patient‐level financial billing information, we performed a series of assumptions regarding the cost of the procedure (Supporting Information: Table S2). All ED visits were charged as level three, hospital admissions costs were based on a three‐night stay, and clinic visits were billed as level four. ED visits were only counted if related to a patient's diagnosis of nephrolithiasis which was determined on a case‐by‐case basis. All CT scans of the abdomen/pelvis were included. Procedures including stent placement or change, and nephrostomy tube placement or exchange, were based on a sum of the hospital‐specific facility fees and professional fees based on the Medicare Fee Schedule. Because facility fees for the hospital vary depending on time spent performing the procedure, it was assumed that stent and nephrostomy tube procedures would use the lowest possible facility fee (<90 min in the OR for stent, <30 min in interventional radiology (IR) suite for nephrostomy tube). The study focused on preoperative financial burden and therefore the cost of the PCNL itself was not included. An aggregate estimate of hospital charges to the patient was determined by a tally of the number of visits, lab tests, imaging, and procedures multiplied by their respective charges as described above.
Stent versus percutaneous nephrostomy (PCN) tube placement was determined by method of upper tract decompression, if necessary, for example, due to infection, renal dysfunction, or intractable pain. This did not include preoperative PCN placement prior to PCNL for surgical access. Stent and PCN counts recorded via available electronic health record (EHR) documentation which included outside facility data, when possible, as long as within the window of diagnosis date and OR date. Encrusted stents were not counted as part of stent placement, primarily because all encrusted stents were already present at time of diagnosis of nephrolithiasis requiring PCNL.
Procedural intervention count was the sum of all stent placement/exchanges and PCN placement/exchanges. Costs of these procedures were used in the overall hospital charge analysis as noted above.
2.3 Statistical analyses
A Kolmogorov–Smirnov test of equality was used to determine normality for variables of interest and showed that the data distributions between the three groups were nonparametric. Therefore, Kruskal–Wallis ANOVA was used for determining group differences between continuous variables. Afterwards, multiple ANOVA of clinic visits, hospital admissions, ED visits, CT scans, and the number of procedural interventions was used to confirm group differences. χ2 was used for categorical variables. p < 0.05 was considered statistically significant.
3 RESULTS
Of the 146 PCNLs performed, a total of 132 PCNLs were analyzed after applying exclusion criteria (Figure 1). The median diagnosis‐to‐OR period for all subjects analyzed was 135 days.
Subjects were split into three groups based on diagnosis‐to‐OR period: less than 4 months, 4–6 months, and greater than 6 months. Doing so resulted in 53, 36, and 43 subjects in each group with mean diagnosis‐to‐OR period of 85, 146, and 266 days, respectively.
Table 1 shows patient characteristics stratified by time to surgery. Across all groups, average age was 45.4 years (SD = 13.1), average BMI was 30.9 (SD = 6.8), 72 (54.5%) were female, 83 (62.9%) were Hispanic/Latino, 60 (45.5%) were documented as unfunded in terms of insurance status, 77 (58.3%) had history of stones, 36 (27.3%) had previous ureteroscopy with laser lithotripsy, 9 (6.8%) had previous PCNL. Between groups, the average age between groups ranged from 44.6 to 46.1 (p = 0.867) and mean BMI ranged from 29.8 to 32.5 (p = 0.220). Between 47% and 58% of subjects were female (p = 0.566), between 49% and 75% were Hispanic/Latino (p = 0.131), between 40% and 53% were unfunded (p = 0.625), 55%–67% had history of stones (p = 0.485), 21%–32% had history of ureteroscopy with laser lithotripsy (p = 0.467), and 2%–14% had history of previous PCNL (p = 0.081). Our study observed no differences between increased time to surgery and age, BMI, gender, ethnicity, or insurance type.
Differences in clinical variables contributing to charge cost between each group are described in Table 2. Delays to PCNL surgery resulted in more CT scans (1.5 vs. 1.6 vs. 2.3, p = 0.019), clinic visits (1.3 vs. 1.8 vs. 2.7, p < 0.001), and total number of procedural interventions (1.5 vs. 1.6 vs. 2.2, p = 0.035. Compared to patients with diagnosis‐to‐OR date less than 4 months, those with PCNL performed in 4–6 months and over 6 months had 7% and 36% higher hospital charges respectively ($27 607 vs. $29 416 vs. $37 622, p = 0.018) (Figure 2). Multiple ANOVA of clinic visits, hospital admissions, ED visits, CT scans, and number of procedural interventions also showed significant differences between groups (p < 0.001). Aggregate costs from all subjects indicated that nephrostomy tube placement, CT scans and ED visits, were key contributors to increased hospital charges (Supporting Information: Figure S2).
No differences were seen in time to PCNL between those requiring stent placement, nephrostomy tube, or no decompression (p = 0.405). Estimated cost specific to our single‐center facility showed similar hospital charges for stent placements/exchanges ($11 794) and nephrostomy tube placements ($12 440). However, nephrostomy tube exchanges were found to have lower costs ($4441).
Stent encrustation was present prior to PCNL surgery in 27 patients (20.6%), with 13 patients with encrusted stents in the <4 months group, 5 in the 4–6 months group, and 9 in the >6 months group. The presence of stent encrustation was not associated with increased time to PCNL (p = 0.472).
4 DISCUSSION
To our knowledge, this is the first investigation evaluating the financial effects of increased time to PCNL surgery in a county hospital setting. The number of clinics, CT scans, and procedural interventions increased as time from diagnosis‐to‐OR date increased, and estimated financial costs were expectedly higher. There was 36% higher financial cost for subjects who received surgery more than 6 months after diagnosis compared to those who received surgery less than 4 months from diagnosis.
Our study, which was performed in a single‐center county hospital, reported higher median days from diagnosis to OR compared to existing literature. While Brubaker and associates document PCNL as having a median time of 52 days from diagnosis to surgery[
5], our median time from diagnosis to OR was 135 days. Known factors associated with delay in other populations include financial concerns and illness due to infection[
5,
6]. However, further studies are needed to evaluate the etiology of delays faced by this underserved population.
Bhojani and associates examined the relationship between ureteral stent duration before PCNL and infectious complications, admissions, imaging, and medical costs[
6]. They found that increased duration of ureteral stent placement prior to PCNL surgery is associated with increased inpatient admissions, infectious complications, imaging, and medical costs. The study did not assess patients who underwent nephrostomy tube placement or did not require upper tract decompression prior to surgery. In comparison, our study included those who underwent stent and/or nephrostomy tube placement. Both studies consistently show that delays in PCNL surgery alone are associated with increased cost and medical resource usage.
In a recent study by Pearle and associates, cost analysis of stent versus nephrostomy tube decompression for obstructing ureteral calculi concluded that stent placement was more expensive[
7]. Our county hospital‐specific findings suggest that while initial nephrostomy placement is similar in cost to stent placement, nephrostomy tube exchanges incur less hospital charges compared to stent exchanges. Additionally, 20.6% of our subjects underwent PCNL with diagnosis of stent encrustation, suggesting that stent placement can incur far more costs compared to nephrostomy tube placement if the stent is retained, encrusts, and requires PCNL surgery for removal.
The financial burden of nephrolithiasis in the literature has been primarily focused on “financial toxicity,” or the direct and indirect costs of treatment to the patient and the resulting effect on quality of life, often measured through subjective data such as the COST questionnaire, QoL questionnaires, and self‐reported financial impact[
8–
10]. While kidney stones are known to pose a significant financial burden to patients, research utilizing objective medical costs to assess the financial impact of PCNL surgery is lacking[
11,
12]. Direct financial cost to the patient is often difficult to quantify given variability across hospitals and payor types. Current guidelines on stone disease do not address the optimal time from diagnosis to definitive stone treatment, mainly due to a lack of data on the subject. Our study introduces an alternative approach to cost evaluation by using an aggregate of hospital‐specific charge estimates for the patient.
Our study is limited in that it involves data from a single institution. It is based on retrospective review, and limitations in data availability exist, in part due to limited accessibility to outside electronic medical records. This study also documented notable delays during the COVID‐19 pandemic, which stressed OR availability and staffing across the country. Our demonstration that more procedures and visits were required for those who had delays in surgery suggests that costs during this period have increased. Hopefully, many of these pressures are alleviated with COVID‐19 in the endemic phase.
These data support the inclusion of commentary regarding stone surgery delays in future surgical guidelines, which do not currently establish a clear consensus on the optimal time scale from diagnosis to surgical intervention[
13]. Future research should be aimed at determining the optimal timing of PCNL surgery to prevent increased financial burden and development of complications.
5 CONCLUSION
This retrospective, single‐center cohort study showed that for our underserved population, increased time prior to PCNL surgery greater than 4 months led to significant increases in healthcare utilization, additional procedures, and hospital charges. Changes aimed at reducing the time from diagnosis to PCNL surgery could reduce the financial burden for both patients and the healthcare system.
2024 The Author(s). UroPrecision published by John Wiley & Sons Australia, Ltd on behalf of Higher Education Press.
PDF
(264KB)
