Changing treatments for primary urolithiasis: impact on services and renal preservation in 16 679 patients in Western Australia

Authors


Abstract

Objectives  To examine changing treatments for the primary presentation of urinary lithiasis and their effects on re-admissions, repeat procedures, cumulative hospital use and renal preservation.

Patients and methods  Linked hospital morbidity records were used to identify first-time admissions for renal and ureteric calculi from 1980 to 1997 in the population of Western Australia. The cases were followed to mid-1999 and actuarial methods used to estimate risks of further hospital admissions and procedures, including the loss of a renal moiety.

Results  Between 1980 and 1997 the total rate of inpatient procedures for urinary stones more than doubled, at a time when the rate of first-time hospital admissions increased by only 13% and the conservative management of stones remained constant at ≈ 59%. The predominant procedure for stone management was initially open lithotomy, replaced in the early 1980s by percutaneous nephrolithotomy and soon supplemented by extracorporeal shock wave lithotripsy (ESWL). The changes in technology led to a four-fold increase in procedural re-admissions within 30 days of primary separation. This was a result of repeated, staged or postponed interventions, often involving the use of stents or a second treatment with ESWL. The risk of surgical intervention decreased from 48% to 32%, whilst the cumulative length of stay over the first year decreased from 7.8 to 3.9 days. The risk of kidney loss declined significantly from 2% to <0.1% during the period.

Conclusions  The main reason for more interventions were short-term procedural re-admissions. ESWL reduced the need for invasive procedures and decreased cumulative hospital stay, despite more re-admissions. Renal preservation improved by a factor of 10.

Introduction

Urolithiasis accounts for 16% of all urological admissions and 1–2% of total hospital admissions in western societies [1,2]. Usually affecting people in the prime of life, this disease causes significant morbidity and loss of productivity. The total economic burden of urolithiasis is estimated to be $1.8–2.4 billion annually in the USA [2,3]. Modern technologies have provided less invasive techniques for dealing with urinary calculi. Percutaneous nephrolithotomy (PCNL) followed by ESWL, and now lasertripsy, have made open surgery rare for complex renal stones [4–6].

Access to a particular method of treatment dictates the practice undertaken; ESWL monotherapy is recommended as a first-line intervention for small-volume renal calculi [7]. ESWL is also recommended as first-line therapy for stones of leqslant R: less-than-or-eq, slant1 cm in the proximal ureter and as being equally acceptable as ureteroscopy to treat stones of leqslant R: less-than-or-eq, slant1 cm in the distal ureter [8]. The recommendations appear to be followed in practice, although an exception is the frequent use of stenting with ESWL, especially in the case of larger ureteric calculi [9]. Ureteroscopic lasertripsy is now becoming more popular and needs further evaluation.

That ESWL is used to treat most calculi in the upper urinary tract is not surprising, given that it has consistently out-performed even minimally invasive endourological procedures in clinical trials of cost-effectiveness [10–14]. Moreover, patients prefer to avoid surgery in the treatment of stone disease, with ESWL being perceived by them as the most desirable treatment option [15].

However, at the population level a different picture can emerge concerning healthcare use and outcomes. After the introduction of ESWL in the USA there was a 71% increase in physician costs [16], and in Quebec the costs of treating stone disease were higher when ESWL became available, because the total (non-surgical plus surgical) intervention rates increased by 52% in women and 34% in men [17]. The growing assortment of treatment options for stones has led to complexity in the way that patients interact with the health system, with ancillary, combination and repeat procedures becoming commonplace during an illness episode. Thus changes in procedural interventions for urolithiasis, and their effects on resource use, renal preservation and the chances of remaining stone-free, need to be evaluated at the level of the health system serving an entire population.

The present study used data obtained through the Quality of Surgical Care Project [18] from the Western Australia (WA) Health Services Research Linked Database [19]. This contains population-based information linking all hospital admissions and deaths in a state of 1.8 million people over a period of three decades. We used linked data starting from 1980 to examine trends in the use of different surgical and non-surgical interventions, re-admissions, repeat procedures, cumulative hospital use and renal preservation in patients presenting with renal and ureteric calculi for the first time.

Patients and methods

The WA Linked Database was used to extract all hospital morbidity data and death records of patients who first separated from an acute hospital in WA in 1980–97 with a principal condition of one or more renal or ureteric calculi (ICD-9-CM 592.X [20]). Patients with any previous mention of renal or ureteric calculus on a hospital-separation record in 1970–79 were excluded, as were patients presenting at first admission with calculi in the bladder or urethra, as well as the upper urinary tract. The linked file was date-stamped 22 March 2000.

There were 16 679 patients admitted for the first time for renal or ureteric calculi in the period 1980–97. Of these, 9095 underwent at least one procedural intervention, including open procedures, transurethral clearance via ureteroscopy or its antecedents, PCNL, ESWL, ancillary endourological procedures including insertion of a ureteric stent, and partial or total nephrectomy including ureterectomy (see Appendix for ICPM [21] and ICD-9-CM [20] procedural codes).

PCNL was first used in WA in 1984. The ICPM system, used to code surgical procedures in the hospital morbidity data before 1988, did not distinguish between open lithotomy and PCNL. Thus, for 1984–87, it was impossible to separate the specific rates of open procedures and PCNL. The first lithotripter in WA was commissioned in 1991, and hence the absence of an ICPM code for ESWL before 1988 was of no consequence. However, a further limitation of the coding was that the placement of a ureteric stent could not be distinguished before 1988 from other ancillary endourological procedures, e.g. ‘push-back’ of stones into the renal pelvis.

In all, 4150 patients underwent multiple procedures up to 30 June 1999, allowing a maximum of 19.5 years of follow-up. There were 12 412 hospital re-admissions, including 5402 for recurrence of upper urinary tract calculus. Total or partial nephrectomies were performed on 229 patients and there were 1539 deaths. Actuarial adjustments were made for loss to follow-up of patients who died.

Age-standardized annual admission and procedure rates in the male and female populations of WA in 1981–1997 were calculated using the direct method and the census population in 1996 as the standard set of weights [22].

The cumulative incidence of hospital re-admission, including procedural re-admissions, within 30 days and 1 year of first-time (index) separation for renal or ureteric calculi were estimated using actuarial methods [23]. The risks of any procedural intervention, any surgical procedure and undergoing a total or partial nephrectomy were estimated similarly, except follow-up commenced from the date of index admission, thereby including index procedures as events. The cumulative length of stay within 1 year was the sum of days spent occupying a hospital bed up to but excluding the first anniversary of the index admission date. Cumulative hospital admissions within 1 year was defined as the number of times the patient was admitted to hospital during the same period. In all analyses using cumulative incidence or usage measures, study subjects were censored at their date of death or, for those who remained alive, on 30 June 1999. Differences in cumulative incidence of outcomes between difference subgroups were assessed for statistical significance using the Wilcoxon (Gehan) test [23].

Results

The mean (range) age of the 16 679 patients was 48.3 (1–95) years and 70.3% were males. In 41.9% the calculi were sited in the kidney alone, in 53.7% only the ureter was affected and in the remaining 4.4% of patients the calculi were located in several sites involving both the kidney and ureter. Over a half (57.1%) of patients were admitted as emergency cases.

Trends in the annual rates of first-time hospital admission and total (first and subsequent) admissions involving a procedural intervention for renal or ureteric calculi in WA from 1981 to 1997 are shown in Fig. 1. There was a net increase in the rates of first-time hospital admission of +21.3% in males (from 77.8 to 94.4 per 100 000 person-years, PY) and +11.7% in females (from 33.4 to 37.3 per 100 000 PY). There were very much larger increases in the rates of inpatient procedures (+153.3% in males; +89.7% in females). The increases in procedural interventions started in the late 1980s and remained a feature throughout the 1990s.

Figure 1.

Age-standardized incidence rates of first-time hospital admission and all procedural interventions for renal and ureteric calculi in males and females in WA in 1981–97. Females, all therapeutic procedures, green open circles; females, first-time admissions, light green closed circles; males, all therapeutic procedures, red open squares; males, first-time admissions, light red closed squares.

Trends in use of different procedural interventions

Table 1 shows the trends in the proportion of index cases treated conservatively and the use of different procedural interventions. Despite the increases in population-based procedure rates shown in Fig. 1 , there was a consistent proportion of index cases treated conservatively, with the percentage fluctuating around a value of 59% throughout the entire 18 years of observation.

Table 1.  First-time hospital admissions for renal or ureteric calculi in WA in 1980–97, and separately for renal and ureteric calculi in WA in 1980–83 to 1995–97, showing trends in proportion treated conservatively and proportions treated using different procedural interventions
Year of
separation
No. of
first-time
cases
Conservative
management,
%
Open
procedure,
%
Transurethral
clearance,
%
PCNL
%
ESWL
%
Ancillary
endourological
procedure, %
Total or partial
nephrectomy,
%
  • *

    Figures not available from 1984 to 1987 – no coding specificity.

198058859.930.47.51.01.7
198166251.230.513.33.92.3
198268850.333.613.12.61.6
198380259.027.912.51.91.5
198479661.9NA*9.3NA*2.91.3
198585859.1NA11.9NA2.41.3
198684359.1NA13.4NA2.71.5
198785057.9NA10.2NA3.30.4
1988101562.59.611.214.45.50.8
198995759.58.011.417.74.90.9
199092257.39.510.119.68.00.7
1991100159.46.46.911.39.910.40.5
199299157.06.36.36.417.810.20.8
1993106160.35.76.45.115.211.30.5
1994112061.54.86.14.715.212.90.2
1995110365.74.76.93.49.814.10.3
1996121658.04.88.33.610.921.10.1
1997120656.83.48.54.412.822.40.0
Total1667959.014.59.45.56.08.90.8
Renal only
1980–83115959.434.92.30.63.7
1984–87145954.338.95.11.22.1
1988–90127555.98.53.628.94.51.3
1991–94185154.04.41.910.124.37.70.8
1995–97125157.23.81.67.121.712.10.3
Total699555.917.32.99.210.35.41.6
Ureteric only
1980–83148951.027.319.53.90.3
1984–87177063.019.016.64.20.1
1988–90150463.09.717.07.07.30.5
1991–94210763.07.210.53.97.114.10.2
1995–97208060.94.512.11.65.524.10.0
Total895060.512.614.72.52.911.60.2

However, there were changes in the use of different treatments in patients who underwent one or more procedural interventions during the index admission. The proportion of patients undergoing open procedures decreased from 30.4% in 1980 to 3.4% in 1997. During the late 1980s, PCNL replaced open surgery as the most common intervention, reaching a peak of 19.6% of cases in 1990, before declining to 3–5% after introduction of ESWL. Transurethral clearance of calculi using ureteroscopy or its antecedents was undertaken in 10–13% of patients before 1992, and in 6–8% of patients thereafter. Since 1992, ESWL and the use of ureteric stents and other ancillary endourological procedures have been the predominant modes of therapy, although ESWL was used less frequently in 1995–97 during index admission than during the first few years after its introduction. Placement of ureteric stents accounted for 77.9% of the 1329 ancillary endourological procedures during index admission in 1988–97. The proportion of cases treated by total or partial nephrectomy at index admission decreased throughout the study period, reaching low levels by 1994.

In Table 1, trends in the use of the different procedural interventions are also shown separately for patients with renal calculi only and those with ureteric calculi only. Transurethral clearance and the use of ancillary procedures were favoured more for ureteric calculi, whereas PCNL and ESWL were used more frequently in patients with renal calculi. Total and partial nephrectomy also occurred more frequently in the surgical management of renal calculi. Otherwise the time trends in the relative use of different procedures were the same regardless of the site of the calculus.

Trends in re-admission for recurrence of calculus and repeat procedures

The risk of hospital re-admission increased markedly, especially from 1991 when ESWL was introduced (Fig. 2); the absolute increases in cumulative incidence of re-admission within the first 30 days from index separation, comparing the risks in 1997 with those in 1980, were + 12.6% for procedural re-admission and + 12.2% for all re-admissions for calculus. At 1 year after the index separation the absolute increases were + 15.0% for procedural re-admissions and + 16.2% for all re-admissions for calculus. These results indicated that all of the increase in re-admissions for calculus within the first 30 days, and four-fifths of the increase with the first year, were explained by a four-fold increase in risk of procedural re-admissions within 30 days, from 4.2% in 1980 to 16.8% in 1997.

Figure 2.

The cumulative incidence of a hospital re-admission, including a procedural re-admission, within 30 days and 1 year of index separation for renal or ureteric calculi in WA in 1980–97: 30-day risk of re-admission for calculus, green open circles; 30-day risk of procedural re-admission, light green closed circles; 1-year risk of re-admission for calculus, red open squares; 1-year risk of procedural re-admission, light red closed squares.

Table 2 shows the patterns of urological practice accounting for the high risk of short-term procedural re-admission, depicting the relative frequencies of different combinations of index and subsequent procedures in patients re-admitted for a procedure within 30 days. Often, patients were treated conservatively during the index admission and re-admitted for ESWL (15.4%) or other interventions. A common scenario was the placement of a ureteric stent during index admission, followed by a second admission for ESWL (19.6%) or other procedure. The third most common pattern, occurring in 7.3%, consisted of ESWL at the index admission followed by a re-admission for a second ESWL. However, the overall pattern was one of considerable diversity in the use of different combinations of therapeutic procedures within the first 30 days of presentation.

Table 2.  Proportions (%) of procedural re-admissions within 30 days of an index separation for renal or ureteric calculi in Western Australia in 1991–97 according to combinations of index and subsequent procedures ( n =928)
Index
procedure
Procedural readmission within 30 days
OpenTransurethral
clearance
PCNLESWLUreteric
stent
Other ancillary
procedure
Total or partial
nephrectomy
None4.612.75.115.413.53.40.1
Open procedure0.10.40.11.40.8
Transurethral clearance0.20.90.51.20.90.6
PCNL0.50.50.40.61.10.10.1
ESWL1.41.61.97.31.70.5
Ureteric stent1.95.11.319.64.62.6
Other ancillary procedure0.20.40.10.50.10.8
Total or partial nephrectomy0.1

Trends in the risk of surgical intervention and hospital use

Despite the large increase in short-term re-admissions for additional procedures, there was a marked decrease in the proportion of patients treated surgically within the first year from initial presentation at index admission. Figure 3 shows that whilst the cumulative incidence of receiving at least one procedural intervention within a year of presentation increased from 47% to 54%, the risk of undergoing a surgical procedure in the first year decreased from 48% to 30%. Most of the reduction in surgical interventions was first observed in patients presenting in 1991, coinciding with the introduction of ESWL.

Figure 3.

The cumulative incidence of any procedural intervention (green open circles), including surgical procedures (red open squares), within a year of index admission for renal or ureteric calculi in WA in 1980–97.

There was a decline in the mean length of hospital stay during index admission, from 6.5 to 2.7 days, and the cumulative length of stay for renal or ureteric calculi summed over the first year from presentation, from 7.8 to 3.9 days (Fig. 4). These trends were not apparently driven solely by the introduction of ESWL in 1991, but occurred smoothly during most of the 18 years of study. Also shown in Fig. 4 are the concomitant increases in the cumulative numbers of total and procedural hospital admissions within a year of presentation, as would be expected from the increase in risk of hospital readmission.

Figure 4.

The mean length of index stay (green open circles), cumulative length of stay (light green closed circles), total (red open squares) and procedural hospital admissions (light red closed squares) within a year of index admission for renal or ureteric calculi in WA in 1980–97.

Trends in renal preservation

Figure 5 shows the trends in cumulative incidence of total or partial nephrectomy within a year of index admission for renal or ureteric calculi. Taking the period averages shown in the figure, there was a substantial decrease in the risk of kidney loss in 1995–97 compared with previous periods ( P =0.007 comparing 0.48% in 1995–97 with 1.11% in 1991–94).

Figure 5.

The cumulative incidence of total or partial nephrectomy (green closed circles) within a year of index admission for renal or ureteric calculi in WA 1980–97, with the period mean (95% CI) shown as the red bars.

Discussion

In this study the rate of hospital admission for procedural interventions for renal and ureteric calculi more than doubled in the population from 1980 and 1997. Within this growing area of healthcare, there were marked changes in methods of treatment used with each new wave of technology. In 1980, open lithotomy accounted for 75% of first-line interventions, but by 1997 it had become the least common intervention. It was replaced initially by PCNL, introduced in 1984, and then by ESWL, introduced in 1991, which dominated treatment patterns by 1992. The introduction of ESWL appeared to be the principal factor associated with rising population-based intervention rates.

The increase in the rate of all interventions for urolithiasis in WA was consistent with trends reported from the USA and Canada [16,17], but the underlying reasons for the present results appeared to be different. Unlike the situation in Quebec [17], there was little evidence of an increase in the number of people being actively treated because of a broadening of criteria [24]. Between 1980 and 1997 there was only a 1% per year increase in the rate of first-time hospital admission for renal and ureteric calculi, suggesting a stable incidence rate of urolithiasis in the population. Of those presenting with the problem for the first time, the proportion treated conservatively during their first hospital admission remained constant, at ≈ 59%. Using record linkage, it was possible to take a longitudinal view of the risk of a new patient receiving any form of intervention during the first year after presentation, and this increased only marginally from 1980 to 1997, from 47% to 54%.

The reason why the population-based intervention rate for upper urinary tract urolithiasis more than doubled in WA was the large increase in procedural re-admissions in the same patient, and usually for the same episode of illness. There was a four-fold increase in procedural re-admissions within 30 days of index admission, and these short-term re-admissions for repeated, staged or postponed interventions accounted for >80% of all re-admissions during the first year of follow-up. The patterns of urological practice responsible for re-admissions within 30 days were complex, involving instances of variable treatment combinations. Commonly, patients underwent initial stabilization with conservative management or placement of a ureteric stent, followed by ESWL. Other patients were readmitted for a second treatment with ESWL or to have their stent removed.

In a series of patients with solitary renal stones, Low et al.[25] found that stent placement made no difference to stone-free rates or the need for a repeat treatment with ESWL. Similarly, Sulaiman et al.[26] reported no effect of ureteric stents on the incidence of steinstrasse in patients with stones of <2 cm, although in patients with larger stones, steinstrasse developed more than twice as often when no stent was placed. In a randomized clinical trial of 400 patients with stones of 1.5–3.5 cm in diameter, Al-Awadi et al.[27] found that placing a ureteric stent before ESWL reduced the risk of steinstrasse from 13% to 6% (P<0.05), and that the incidence of steinstrasse increased with the size of the calculus. Thus, although the use of ureteric stents in patients undergoing ESWL for urinary calculi remains controversial, there is an empirical basis for stent placement in the management of larger stones. Hollowell et al.[9], in their Internet and postal survey of endourological practice patterns among American urologists, found a pattern of practice consistent with this evidence. Placement of stents was reported by 25% of urologists for a 1.0 cm calculus, 57% for a 1.5 cm calculus and 87% stated they would place a stent for a calculus 2.0 cm in diameter.

The greater need for several procedures in patients treated with ESWL than with other therapeutic methods was documented by the Clinical Guidelines Panel from results of their meta-analyses of 110 and 327 studies [7,8]. For staghorn renal calculi treated with ESWL monotherapy, the mean number of ESWL procedures was 2.12, compared with 1.49 per patient treated with PCNL monotherapy [7]. For calculi in the distal ureter, there was a mean of 1.24 ESWL treatments needed, compared with 1.01 for ureteroscopy [8]. The present results show the effect of these different ratios on the healthcare of a population.

The shortcomings of the present study should be considered. There were no data available on ambulatory care and it was possible that several episodes of renal colic did not require the patient to be hospitalized. However, the study identified all instances of the relevant therapeutic procedures, as these were not delivered on an outpatient basis in WA during the study period. The study was based on administrative health data and provided no details of the type that would be expected in smaller clinical studies. Because of the coding limitations, we could not distinguish between calculi presenting in the proximal and distal ureter, nor was information available on stone diameter, shape or laterality. Despite its limitations, the strengths of the study were that it was population-based, covered a period of 17 years, identified virtually all of the initial and subsequent therapeutic procedures, and was precise, because there were many patients.

In conclusion, the changes in urological practice in the management of urinary calculi appear to have conferred at least three benefits on the population and the health system. The introduction of ESWL has greatly reduced the chance of a patient undergoing any form of surgical intervention, either at initial presentation or during subsequent follow-up. The risk of an invasive procedure within a year of index admission decreased by 38%. A second advantage has been a much shorter hospital stay, both during the first admission and even when stays were summed over a year, despite the increase in short-term procedural re-admissions. Using estimates of cost per bed-day for conditions of the kidney and urinary tract in Australian hospitals [28], the decrease from 7.8 to 3.9 days in the mean cumulative length of stay within the first year of treatment from 1980 to 1997 (Fig. 4) corresponded to a reduction from Aus$ 4748 to Aus$ 2374 per patient, albeit that such values do not consider changes in resource use per inpatient-day. These outcomes are generally consistent with results of other population-based research [16,17] and clinical trials of the cost-effectiveness of the noninvasive and minimally invasive treatment methods [10–14]. The third benefit has been the increased chance of renal preservation. The risk of losing a kidney to urolithiasis within the first year after presentation has decreased in WA from 1–2% in the early 1980s to a risk of <0.1%. Renal preservation values after treatment for urolithiasis are rarely reported; the Nephrolithiasis Clinical Guidelines Panel reported no data on kidney loss after ESWL, but reported average risks of 3.8% after an open procedure and 1.6% after PCNL for staghorn calculi [7]. Kidney loss in patients with ureteric calculi was not reported [8]. This is surprising, given that renal preservation is a highly desirable outcome and even a risk of 0.1% of kidney loss is important. It should be considered as one of several key indicators of the quality of surgical care for urolithiasis at the population level.

Acknowledgements

This research was supported by the National Health and Medical Research Council of Australia. The Health Services Research Linked Database was supported by the Lotteries Commission of Western Australia. We are indebted to Professor David Fletcher, Mr Michael Lawrence-Brown and fellows of the WA Branch of the Royal Australasian College of Surgeons for their collaboration in the Quality of Surgical Care Project.

C. D'Arcy J. Holman, Centre for Health Services Research, Department of Public Health, The University of Western Australia, Nedlands, Western Australia 6907.
e-mail: darcy@dph.uwa.edu.au

Appendix

The procedural codes used is this study were as follows: open procedures (ICPM 1–651, 5–550, 5–551, 5–562 [21]; ICD-9-CM 55.01, 55.02, 55.1X, 55.21, 55.22, 56.2 [20]); transurethral clearance via ureteroscopy or its antecedents (ICPM 5–560; ICD-9-CM 56.0); PCN (ICD-9-CM 55.03, 55.04), ESWL (ICD-9-CM 98.51); ancillary endourological procedures including insertion of a ureteric stent (ICPM 1–650, 1–654, 5–561; ICD-9-CM 56.1, 56.31, 59.8); and partial or total nephrectomy, including ureterectomy (ICPM 5–552, 5–553, 5–554, 5–563; ICD-9-CM 55.3X, 55.4, 55.51, 55.52, 56.4X).

Ancillary