Primary prevention of nephrolithiasis is cost-effective for a national healthcare system

Authors


Yair Lotan, Department of Urology, The University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd., Dallas, TX 75390, USA. e-mail: Yair.lotan@utsouthwestern.edu

Abstract

Study Type – Therapy (cost-effectiveness meeting)

Level of Evidence 2b

What's known on the subject? and What does the study add?

One of the major problems with nephrolithiasis is the high rate of recurrence, which can effect up to 50% of patients over a 5-year period. Patients with recurrent stones are recommended to increase fluid intake based on prospective studies that show a reduction in recurrence rates in patients who intake a high volume of water. Strategies to reduce stones in recurrent stone formers are quite effective with a >50% risk reduction with increased fluid intake alone. Unfortunately, despite a high societal cost and morbidity, there are no prospective studies evaluating the benefit of fluid intake to prevent stone disease in subjects without a prior history of stone but at risk for stones.

The budget impact analyses show that prevention of nephrolithiasis can have a significant cost savings for a payer in a healthcare system and reduce the stone burden significantly. Future studies will need to assess the feasibility and effectiveness of such an approach in a population.

OBJECTIVE

  • • To evaluate the impact of primary prevention of stones using a strategy of increased fluid intake.

SUBJECTS AND METHODS

  • • A Markov model was constructed and analysed using Excel to calculate and compare the costs and outcomes for a virtual cohort of subjects with low vs high water intake.
  • • A literature search was used to formulate assumptions for the model including an annual incidence of urolithiasis of 0.032%, annual risk of stone recurrence of 14.4% and 40% risk reduction in subjects with high water intake.
  • • Costs were based on resource utilisation from the Delphi panel and official price lists in France.
  • • Outcomes were based on payer perspective and included direct and indirect costs and loss of work.

RESULTS

  • • The base-case analysis found total cost of urolithiasis is €4267 with direct costs of €2767, including cost of treatment and complications. The annual budget impact for stone disease based on 65 million inhabitants is €590 million for the payer.
  • • The use of high water intake by 100% of the population results in annual cost savings of €273 million and 9265 fewer stones. Even if only 25% of the population is compliant, there is still a cost saving of €68 million and 2316 stones.
  • • The model was evaluated to determine the impact of varying the assumptions by ±10%. For example, when the incidence of stone disease is increased or decreased by 10% then the mean (range) baseline cost will change by €59 (531–649) million for the payer and savings will either increase or decrease by €27 (246–300) million.
  • • The largest impact on cost savings occurs when varying risk reduction of water by 10% resulting in either a mean (range) increase or decrease by €35 (238–308) million.
  • • Varying cost of stone management by 10% has an impact of ±€17 million. Varying other factors such as stone recurrence by 10% has only an impact of ±€9 million and varying risk of chronic kidney disease ±€1 million, as they affect only a portion of the population.

CONCLUSIONS

  • • The budget impact analyses show that prevention of nephrolithiasis can have a significant cost savings for a payer in a healthcare system and reduce the stone burden significantly.
  • • Future studies will need to assess the feasibility and effectiveness of such an approach in a population.
Abbreviations
CKD

chronic kidney disease

ESRD

end-stage renal disease

REIN

Réseau Epidémiologie et Information en Néphrologie.

INTRODUCTION

Nephrolithiasis is a worldwide problem that can result in significant morbidity and cost. The overall prevalence of stone disease varies based on age, gender, race, and geographic location with reported rates of 1–5% in Asia, 5–9% in Europe, 13% in North America, 4% in Argentina, 5.5% in Mexico and 20% in Saudi Arabia [1–3].

The consequences of stone disease can range from acute episodes that result in pain, loss of work, need for hospitalisation and surgery to chronic conditions such as renal insufficiency. Stone-related events frequent working-age adults more often than children or the elderly compounding their financial impact [4]. While the worldwide economic impact is unknown, the total annual medical expenditures for nephrolithiasis in the USA were >$2.1 billion in the 2000 alone [5]. European countries also have a considerable prevalence and cost for managing stone disease [6].

One of the major problems with nephrolithiasis is the high rate of recurrence, which can impact up to 50% of patients over a 5-year period [7]. As a consequence, patients with recurrent stones are recommended to increase fluid intake based on prospective studies that show a reduction in recurrence rates in patient who intake a high volume of water [8,9]. Dietary modifications have also been shown to reduce the risk of recurrent stone formation [10,11]. Strategies to reduce stones in recurrent stone formers are quite effective with >50% risk reduction with increased fluid intake alone [8,9]. Unfortunately, despite a high societal cost and morbidity, there are no concerted efforts at preventing stone disease in subjects without a prior history of stone disease. One remote study found that an educational campaign to increase fluid intake in a village in a desert environment resulted in a 90% risk reduction in stones suggesting the feasibility of such an approach [12]. Furthermore, a recent study on the cost-effectiveness of primary prevention found that primary prevention could be cost-effective if the incidence of stones was >4.3% yearly or the cost of prevention was <$23 per person yearly assuming a 1% stone incidence [13]. Limitations of this analysis included the fact that it applied to only a 1-year analysis. This may be applicable in some scenarios but there are some longer term consequences of stone disease, e.g. the risk of recurrence and potential complications, e.g. chronic kidney disease (CKD), which can take years to manifest but result in significant morbidity and cost. In order for a payer, such as a government or insurance company, to determine the financial benefits of instituting a prevention plan, the long-term implications can be critical.

Budget-impact analyses are used together with cost-effectiveness analyses to estimate the affordability of a new healthcare intervention to specific healthcare-decision makers. Cost-effectiveness analyses focus on a representative individual or cohort and estimates costs and benefits of the new intervention over a time horizon that captures all relevant costs and benefits. By contrast budget-impact analyses estimate the impact of the new intervention on short- or longer-term annual healthcare budgets for a population. These types of financial analyses are performed from the perspective of the payer, including mainly medical costs, and the costing is based on actual prices and tariffs.

Until recently no systematic and specific approach has been developed for the assessment of the impact of food products on health and health-related quality of life in our society. The goal of the present study was to evaluate the impact of primary prevention of stones using a strategy of increased fluid intake, using the French healthcare system as an example. This health economic assessment of increased water intake is performed taking into consideration a recent policy paper on health economic evaluations in relation to nutrition [14].

PATIENTS AND METHODS

A decision analytic model was developed to estimate the cost-effectiveness of high water intake (2 L/day) vs low water intake (<2 L/day) from the perspective of the French payer and society in 2010. The analysis was performed for a hypothetical cohort of persons. Data sources used included published literature, clinical trials, official French price/tariff lists and national population statistics. A French expert validated the research plan for the model (structure, assumptions and data sources) taking into consideration the specific French healthcare setting.

MODEL DESIGN

A Markov model was constructed and analysed using Excel to calculate and compare the costs and outcomes for a virtual cohort of subjects with low vs high water intake. Markov modelling is most appropriate for use in situations where events reoccur over time and persons/patients move among a finite number of health states over the period being considered. A Markov process model describes several discrete states of health in which a person can be at time t = n, as well as the states of health into which the person may move at time t = n + 1. The progression from t = n to t = n + 1 is called a ‘cycle’. Markov modelling allows one to account for subjects as their health condition changes over time. All clinically important events are modelled as transitions from one state to another state. Probabilities are associated with each change between two states; these are termed ‘transition probabilities’. Each transition probability is a function of the state of health and the treatment. Our Markov model is shown in Fig. 1. The structure for the submodel ‘low water intake’ is similar to that of ‘high water intake’, which has been shown in Fig. 1. A 1-year cycle was used in the Markov model and the follow-up time was 5 years to life time. Mortality was also incorporated in the model by definition of the health state ‘dead’. The Markov states are:

Figure 1.

Markov model for increased water intake.

  • • No nephrolithiasis
  • • No nephrolithiasis after nephrolithiasis
  • • CKD
  • • dead

The transitions between health states during each cycle may be categorised into:

  • • No nephrolithiasis: Persons may develop nephrolithiasis or may not develop nephrolithiasis. A person who does not develop nephrolithiasis remains in the state ‘no nephrolithiasis’. Persons with lithiasis will move to the state ‘No nephrolithiasis after nephrolithiasis’. An episode of nephrolithiasis will in a proportion of patients lead to complications (pyelonephritis), which are included in the model as ‘events’.
  • • A person who does develop nephrolithiasis may develop end-stage CKD (dialysis, transplantation).

STUDY POPULATION

The analysis was performed for a cohort of healthy persons drinking <2 L fluid/day, which represent ≈80% of the French population [15].

TIME HORIZON

The base-case analysis for the health economic analysis was based on a 25-year follow-up period in order to capture the impact of increased water intake on long-term morbidity and mortality resulting from complications of nephrolithiasis. The budget-impact analysis is based on a 5-year period.

PERSPECTIVE AND SETTING

The perspective of the study in the base-case analysis was that of the French payer including direct medical costs and also indirect costs due to lost productivity.

COST ASSESSMENT

The cost assessment is based on the assignment of costs to the health states. The costs of each health state are determined by the resource use associated with a health state. We focused on the costs associated with nephrolithiasis, treatment related complications and CKD. The model was designed to assess absolute difference in stone-related events such that all other health care use was excluded in the model. The incremental cost difference is based on: (i) a difference in costs associated with nephrolithiasis and its complications (acute and long term), and (ii) a difference in costs associated with high and low water intake. The base-case analysis was based on 3% discount rate for economic outcomes and clinical outcomes. The model cost inputs are taken from different years due to the availability of data. All have been corrected to 2009 costs as no 2010 inflation data were available at the time of the present study.

ANALYSIS

Base-case analysis

Cost-effectiveness analysis compares the costs and outcomes for alternative therapeutic strategies. An incremental cost-effectiveness ratio represents the additional cost and effectiveness obtained, when increased water intake is compared with standard therapy.

Clinical event probabilities

Table 1 [9,16–18] provides a summary of all clinical input variables.

Table 1. Clinical probabilities in the model
ProbabilitiesValueSource
  1. Tx, transplantation.

Incidence nephrolithiasis (normal population)0.000323REIN report [17]
CKD due to nephrolithiasis0.008REIN report [17]
Risk reduction with high water intake:  
• primary prevention0.626Curhan et al.[16]
• secondary prevention0.333Borghi et al.[9]
Mortality:  
• normal population0.0124 http://www.indexmundi.com/france/life_expectancy_at_birth.html
• dialysis0.185Nuijten et al.[18]
• Tx0.040
Transitions in CKD:  
• dialysis to Tx0.15Nuijten et al.[18]
• Tx to dialysis0.05

Probability nephrolithiasis:

The annual incidence of nephrolithiasis in the general population of France was estimated at 21 000 leading to annual incidence of 0.032% based on a French population of 65 million. This figure is in agreement with the literature. A study by Frank and De Vries [12] reports an incidence of 0.07% and 0.80% for two different regions in Israel.

The risk of stone recurrence in patients with prior stones was based on a French study that reported a 14.4% annual recurrence rate [19]. This is similar to a report by Borghi et al.[9] that found a risk of recurrence of 37.8% over 3 years, which is annual risk of 12.6%.

The assumption is that the incidence of nephrolithiasis corresponds with 80% of the current population taking <2 L fluid/day and the other 20% taking ≥2 L/day.

Risk reduction of stones based on high water intake

We assumed a 40% risk reduction based on a report by Curhan et al.[16] that found that the annual stone incidence was 0.23% in subjects without high water intake and 0.14% after increasing water intake. This is supported by a randomised prospective study in recurrent stone formers that found a 55% risk reduction based on high intake of water (12.1% with and 27.0% without high water intake) [9].

Nephrolithiasis and risk of end-stage CKD

The Réseau Epidémiologie et Information en Néphrologie (REIN) report found that 0.8% of end-stage renal failure results from nephrolithiasis [17]. The prevalence of end-stage CKD is ≈70 000 and annual incidence is ≈15% (10 500) [20]. The annual incidence of CKD due to nephrolithiasis is 0.8% of 10 500 (n= 84). Thus the annual risk of renal failure due to nephrolithiasis is 0.12% (84 cases of lithiasis in 70 000 cases of end-stage CKD).

Treatment costs

The costing was based on resource use from the Delphi panel and official price lists in France. Resource use was defined by means of a modified Delphi panel consisting of five GPs, four urologists and one nephrologist to reflect the treatment patterns in France.

Direct costs included inpatient and outpatient care. Transportation costs were not included. Hospitalisation costs (nursing, care, pharmacy, diagnostic tests, laboratory tests, staff, general equipment, administration, security, central supply, dietetics, social services, etc.) were assessed using the French Diagnosis-Related Groups that fund all costs (fixed and variable). Outpatient care and medical procedures performed outside the hospital were valued according to the Social Health Insurance tariffs. Unit costs were estimated for 2010 and expressed in Euros (€). Indirect costs are calculated using an average number of days of work lost due to disease [21,22] The data concerning the number of workdays lost were defined by the Delphi panel. The valuation of days out of work is based on income per capita per day, which was adjusted for the working population and for 250 days of work per year (cost per day was €300; Table 2) [18,23].

Table 2. Costing applied in the model
Item Costs, €Source
  1. GP, general practitioner; IV, intravenous; CKD, chronic kidney disease; HD, hemodialysis; PD, peritoneal dialysis; Tx, transplantation.

ConsultationUrology23.00Source: http://www.ameli.fr/professionnels-de-sante/medecins/index.php
Gynaecology23.00
GP22.00
Cardiologist45.73
ProceduresCystoscopy38.40
Shockwave lithotripsy250.80
Ureteroscopy and laser lithotripsy213.39
Ureteric stent placement1399.16
Diagnostic Procedures Cost/item
Urine culture18.9
Microscopy urine4.05
Laboratory blood19.98
Radiography19.95
Pyelogram i.v.85.12
Renal function (creatinine blood)10.80
 Annual costs
CKDCKD – HD54 463Nuijten et al.[18]
CKD – PD61 862
CKD – Tx (year 1)4357
CKD – Tx (other years)1435
Complications/hospitalisation  Source: http://stats.atih.sante.fr/mco/catalogmco.php
NephrolithiasisCost/eventLength of stay, days
889.671.38
2271.755.08 (Delphi 3.9)
3313.527.68
Fever4124.9711.81
1587.203.03 (Delphi 2.6)
3596.148.38
4681.1311.50
Pyelonephritis6062.4913.92
1587.203.03 (Delphi 2.0)
3596.148.38
4681.1311.50
Renal failure6062.4913.92
1828.533.94 (Delphi 1.6)
3711.968.15
5289.9911.68
€6580.1216.12
Number of days off  
Indirect costs Unit cost per dayCost/day [23]
  €300.00 (Delphi 7.3 days)

Costs

The base-case analysis indicated that the total cost of nephrolithiasis was €4267 with direct costs of €2767, including the cost of treatment of nephrolithiasis and complications. The total cost for stone management was similar to previously published rates [6].

The model assumed that 33% of patients were initially diagnosed and treated by a GP. Referral to specialist was subsequent to a visit to the GP for a proportion of patients, who were not successfully treated and 67% of patients were directly diagnosed and treated by specialists.

RESULTS

BUDGET IMPACT

The annual budget impact for stone disease based on 65 million inhabitants was €590 million for the payer. The use of high water intake by 100% of the population would lead to annual cost savings of €273 million and 9265 stones (Table 3). Compliance with increased fluid intake may vary considerably and the model was used to evaluate scenarios with reduced levels of compliance. Even if only 25% of the population was compliant, there was still a cost saving of €68 million and 2316 stones (Table 3).

Table 3. Cost of care results (annual budget)
Population compliance with 2 L fluid/daySavings (€ million)Number of stone events prevented
Medical costsIndirect costsTotal
100%177962739265
80%141772187412
75%132722056949
50%88481374633
25%4424682316

The model was evaluated to determine the impact of varying the assumptions by ±10%. For example, when the incidence of stone disease is increased or decreased by 10% then the mean (range) baseline cost will change by €59 (531–649) million for the payer and savings will either increase or decrease by €27 (246–300) million. The largest impact on cost savings occurred when varying risk reduction of water by 10% resulting in either an increase or decrease by €35 (238–308) million. Varying cost of stone management by 10% had an impact of ±€ 17 million. Varying other factors, such as stone recurrence by 10% had only an impact of ±€9 million and varying risk of CKD ±€1 million, as they affect only a portion of the population.

COST-EFFECTIVENESS

The cost-effectiveness analysis was from the perspective of an individual in society. The base-case analysis is based on a follow-up of 25 years and found that increasing water intake can save from a payer perspective €30 on average per person (Table 4).

Table 4. Cost-effectiveness results for individuals with varying compliance of water intake over 25 years
 Costs (€)Stone events
100% uptake  
 High water intake300.009
 Low water intake650.012
 Difference−35−0.003
80% uptake  
 High water intake370.010
 Low water intake650.012
 Difference−28−0.002
75% uptake  
 High water intake380.010
 Low water intake650.012
 Difference−27−0.002
50% uptake  
 High water intake470.011
 Low water intake650.012
 Difference−18−0.001
25% uptake  
 High water intake560.012
 Low water intake650.012
 Difference−90.000

The advantages for high water intake are also present with lower compliance. While the differences between high and low water intake appeared low, it is due to the fact that the stone incidence overall in society is low but represents an average saving for every person in society. When performing sensitivity analyses varying input variables by ±10%, the results showed that incidence of nephrolithiasis, increased water intake, and medical costs were most sensitive variables with variance of ±€4, €3, and €2, respectively. For the other parameters, sensitivity analyses showed a marginal effect on the outcomes.

DISCUSSION

There is considerable concern about rising healthcare costs. Diseases of high prevalence such as nephrolithiasis can result in significant healthcare expenditures. Annual medical expenditures for nephrolithiasis in the USA were >$2.1 billion in the 2000 alone [5] and a recent French study found that nephrolithiasis resulted in 144 324 admissions in 2009 at a total cost of >€168 million [24]. One way to reduce healthcare costs is prevention of common diseases. This type of approach is broadly used by society with strategies, e.g. smoking cessation, encouraging exercise and healthy diet in schools, and vaccination, to name a few. The main factors in considering whether to institute a prevention strategy rests on identifying diseases of sufficient prevalence, morbidity and cost as well as strategies that are cost-effective. Nephrolithiasis is both a common problem affecting 5–9% of the population in Europe and 13% in North America [1] and one of considerable cost and morbidity. Several studies have found that increased water intake can reduce the risk of stone disease by 40–50% [9,12,16].

The present study evaluated the impact of high water intake using a Markov model, which took into consideration the cost and incidence of stone disease as well as risk reduction from increased water intake. The study was performed from the perspective of the payer, which in this case represents either insurance companies or the government, and results were given as an annual budget-impact analysis or cost-effectiveness analysis from an individual's perspective. In either case, the increased intake of water results in considerable cost savings ranging from €273 million in a highly compliant population to €68 million in a population of low compliance (25%) based on the French healthcare system as a model. Such a strategy can also have the benefit of reducing the incidence of stones and associated morbidity and complications. The cost savings in an annual budget underestimate the true savings that can occur over time as there is reduction in long-term consequences, e.g. CKD, which impact a small number of patients but at considerable cost.

The present study has advantages over a recent cost-effectiveness study on prevention in the use of a Markov model, which allows evaluation over time taking into consideration future stone recurrences and potential complications, e.g. infections and CKD, as well as death from other causes [13]. One major assumption, is that the cost saving was calculated based on the perspective of French public healthcare funds, and consequently cost burden for the population to purchase the water were not taken into account. Another consideration in the present study was that the stone incidence of 0.03% was taken from the entire population [17]. The distribution of stone occurrence is not equal for all groups in the population. It is higher in working age adults and there are risk factors, e.g. geography, occupation, obesity, gender and ethnicity, which can impact the incidence of stones. Concentrating prevention on portions of the population at higher risk will significantly increase the cost effectiveness.

There are potential limitations with the present analysis. First, it is based on costs in France and these costs vary between countries. However, the model can be applied using costs and stone prevalence for any country. For example, the stone prevalence in the USA is higher than in Europe and cost saving would be higher for similar risk reduction and compliance. The model also depends on the availability of clean drinking water. If there is no clean water then there would be added costs to provide water. Furthermore, there could be additional costs to provide education to a population to improve compliance with fluid intake. The purpose of the present study was to evaluate the potential cost-effectiveness of a primary prevention programme and not the implementation. However, there have been programmes to impact public behaviour. France has a French nutrition and health programme (Le Programme National Nutrition Santé) that spent a total of €10 million to promote ‘Fruit and Vegetables’ in 2006 [25] Compliance with such a programme is difficult to ascertain and is an important area of future study. The model can be used to determine how much funding could be available for an awareness campaign while still providing cost-benefits.

One further assumption is that fluid intake may prevent end-stage renal disease (ESRD) in stone formers. While this represents a very small proportion of the financial impact, as ESRD accounts for only 0.12%, the assumption needs to be proved. However, there is some evidence that fluid intake can delay ESRD in patients with primary hyperoxaluria [26]. Even if the assumption is incorrect, it will have a negligible impact on the financial outcomes of the model. Finally, in rare cases, there may also be harm from very high fluid intake. However, this mainly applies to patients with CKDs, hepatic cirrhosis or ascites and these populations represent a small part of the total adult population and generally have a medical follow-up. Most ‘normal’ subjects can tolerate even a prolonged high fluid intake without any impact on plasma osmolality [27].

In conclusion, the budget-impact and cost-effectiveness analyses show that prevention of nephrolithiasis by drinking an adequate amount of water can have a significant cost-savings for a payer in a healthcare system and reduce the stone burden significantly. Future studies will need to assess the feasibility and effectiveness of such an approach in a population.

CONFLICT OF INTEREST

Yair Lotan, Michel Daudon, Laurent Molinier, Ivan Tack and Mark J. C. Nuijten served as Consultants for Danone Inc. Inmaculada Buendia Jiménez and Irene Lenoir-Wijnkoop are Employees of Danone Inc. Source of funding: Danone.

Ancillary