Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Apr 2.
Published in final edited form as: Lancet. 2026 Mar 21;407(10534):1171–1181. doi: 10.1016/S0140-6736(25)02637-6

Prevention of Urinary Stones with Hydration: A Randomised Clinical Trial of an Adherence Intervention

Alana C Desai 1,†, Naim M Maalouf 2,‡, Jonathan D Harper 3,‡, Sri Sivalingam 4, John C Lieske 5, H Henry Lai 1,†, Peter P Reese 6, Hunter Wessells 3, Hongqiu Yang 7, Hussein R Al-Khalidi 7,8, Ziya Kirkali 9, Gregory E Tasian 10,*, Charles D Scales Jr 7,11,*, for the Urinary Stone Disease Network Investigators
PMCID: PMC13043023  NIHMSID: NIHMS2159793  PMID: 41864748

Summary

Background:

Increased fluid intake is universally recommended to decrease the risk of recurrent urinary stones; however, adherence is challenging. The effectiveness of interventions to maintain high fluid intake has not been well studied. We sought to determine whether a multi-component behavioural intervention program to promote high fluid intake reduces symptomatic stone recurrence, compared with control.

Methods:

Participants aged ≥12 years with a history of urinary stone disease and low 24-hour urine volumes based on current guidelines were enrolled at 6 academic medical centers in the United States from 2017 until 2024. Participants were randomised in a 1:1 ratio to a multi-component behavioural intervention designed to promote increased fluid intake or to the control arm receiving guideline-concordant care. The primary outcome was symptomatic stone recurrence defined as stone passage or procedural intervention for stone(s) during a 2-year follow-up period. Secondary outcomes included 24-hour urine volume, urinary symptoms, and radiographic stone recurrence or growth; hyponatremia requiring hospitalization was the safety endpoint.

Findings:

1658 participants were randomised to intervention (n=826) and control (n=832) groups (median age 44 years; 946 [57.1%] female). Among these 1658 participants, 1103 (66.5%) were recurrent stone formers. Symptomatic stone events occurred in 154 (18.6%) participants in the intervention group and 165 (19.8%) in the control group (hazard ratio 0.96; 95% confidence interval [CI] 0.77 to 1.20). No episodes of hyponatremia requiring hospitalization (safety endpoint) were reported; asymptomatic hyponatremia was reported in 1.5% of intervention participants, vs 0.2% of controls (p=0.018).

Interpretation:

A behavioural intervention program to promote fluid intake for secondary stone prevention modestly increased urine volume but did not reduce recurrent stone events compared with guideline-based care during a 2-year follow-up period.

Trial Registration:

ClinicalTrials.gov (NCT03244189).

Urinary stone disease is a common disorder of mineral metabolism marked by episodic painful events1 that negatively impact physical, social, and emotional health.2 Guideline-based prevention strategies emphasize high fluid intake as a cornerstone of reducing recurrence risk.4,5A Cochrane review included a single randomised controlled trial that demonstrated increased water intake reduced urinary stone recurrence, rated as low certainty evidence.6 This seminal trial by Borghi,7 comparing high fluid intake versus no additional fluid intake in 199 participants, demonstrated benefit for increased hydration. Thus, achieving and maintaining high fluid intake could provide safe and effective secondary prevention of symptomatic urinary stones. However, maintaining high fluid intake is a formidable challenge: the average increase in urine volume following counseling by physicians is only 300 mL/day.8 Therefore, the PUSH study focused on improving adherence to high fluid intake, testing an intervention grounded in behaviour change theory and focusing on addressing common barriers to high fluid intake.

Barriers to achieving and maintaining high fluid intake are manifold. Among these, patients frequently identify lack of awareness of volume consumed and/or intake goals, forgetting to drink, and perceived need to drink as critical impediments.9 Other barriers include practical concerns such as lack of access to water, bathroom access, or competing time demands. Similar types of barriers to behaviour change exist for many diet-related health conditions, including obesity, for which the United States Preventive Services Task Force recommends multi-component behavioural interventions to improve weight status and reduce the incidence of type 2 diabetes.10 The PUSH study intervention is designed as an adaptive, multi-component intervention leveraging multiple behaviour change techniques (BCTs)11 to promote adherence to fluid intake. As part of a contingency management approach, the PUSH study leverages financial rewards for behaviour change, which is a well-validated and widely accepted way to incentivize a broad range of health behaviours, including reducing substance use, increasing physical activity, and promoting weight loss.12 While contingency management is effective at eliciting behaviour change, it may be more effective when utilized as part of a multi-component behaviour change strategy.13 In the PUSH trial, the primary intervention to support increased water intake was providing a “smart” water bottle [add objects to the environment BCT (behaviour change intervention ontology (BCIO, bciosearch.org):007156); self-monitoring BCT (BCIO:006137)] and loss-framed financial incentives (provide aversive material consequence for behaviour BCT; BCIO:007243) for meeting water intake targets. Additional intervention in the form of health coaching (structured problem-solving—guide how to perform behaviour BCT; BCIO:007050) was available to individuals who consistently failed to meet water intake targets over any two week period in the study, to help overcome practical barriers to fluid intake. This type of adaptive intervention strategy is common in health behaviour change, with additional support provided if a patient fails to meet goals.14 Health coaching and other low-touch approaches are commonly added to behaviour change interventions when people are not meeting their goals.15 The rationale for adding structured problem-solving for participants falling short of their goals was intended to provide assistance for identifying and solving barriers to meeting the prescribed water intake.

The PUSH study selected recurrent symptomatic urinary stones as the primary endpoint, rather than the more proximate outcome of increased fluid intake (or 24-hour urine output, which is the guideline-based clinical surrogate). The efficacy of behavioural interventions should ideally be determined by clinically important outcomes that are meaningful to patients, because surrogate outcomes may overestimate the benefits of the intervention.14 PUSH tested the hypothesis that this adaptive, multi-component behavioural intervention to promote fluid intake would be more efficacious to 1) increase urine volume and 2) reduce recurrent symptomatic stone events than a control arm receiving guideline-based care from their usual physicians.

METHODS

Trial design and oversight

The PUSH trial was an investigator-initiated, randomised controlled trial conducted between October 4, 2017 and April 16, 2024, at 6 medical centers in 5 states that participate in the Urinary Stone Disease Research Network (USDRN). The rationale and design of the PUSH trial have been published previously.16 The protocol was approved by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) and institutional review boards of the Scientific Data Research Center and participating institutions. Patients and members of the public reviewed and provided input on the protocol. Protocol amendments and rationale are detailed in the Supplement. Adult participants provided written informed consent and adolescents provided verbal assent along with consent by legally authorised representatives. A data safety monitoring board convened by the NIDDK met regularly to assess trial progress and participant safety. Ethical approval was provided by the Duke University Health System Institutional Review Board (Pro0008327).

Participants

English-speaking patients aged ≥12 years with a recent symptomatic urinary stone event were screened. Patients receiving care at USDRN clinical centers and at outside institutions were eligible. Screening, recruitment, and enrollment occurred primarily in connection with clinical care for urinary stone disease. During and after the COVID-19 pandemic, participants could also be screened, recruited, and enrolled remotely. Recent symptomatic stone events were defined as spontaneous stone passage or receiving a procedural intervention to remove stones within the prior three years, or a symptomatic stone event within five years if new stone(s) were detected on subsequent diagnostic imaging such as ultrasound or computed tomography. Eligible participants were required to have a baseline 24-hour urine volume <2.0 L/day for adults and adolescents aged 12–17 years weighing ≥75 kg, and <25 mL/kg body weight/day for adolescents weighing <75 kg. Eligible participants were required to have access to a mobile device that could sync with a Bluetooth-enabled smart water bottle (eFigure S1).16 Full eligibility criteria are available in the protocol, with the intention enrolling participants with idiopathic stone disease (i.e., not monogenic or from a surgical condition such as bariatric surgery) (Supplemental Material: Trial Protocol).

Randomisation and Masking

Participants were randomised in a one:one ratio to an adaptive multi-component behavioural intervention focused on increasing fluid consumption (eMethods in Supplementary Appendix) or a control arm with a follow-up of 2 years. Randomisation was stratified by age group (adult or adolescent), stone history (first-time or recurrent stone former), and study site. Randomisation assignment was computer-generated remotely at the SDRC. Every participant received a Bluetooth-enabled smart water bottle that measured and recorded fluid intake (eFigure S2). Investigators, treating physicians, outcome assessors, and adjudicators were blinded to group assignment. It was not feasible to blind study participants, coordinators, or health coaches. At enrollment, participants aged ≥18 years underwent a low-dose, non-contrast computed tomography (CT) scan, participants <18 years had a renal ultrasound; all completed a 24-hour urine collection.

The intervention consisted of a fluid prescription (set measurable behaviour goal BCT BCIO:007300), financial incentives to adhere to fluid prescription, health coaching to overcome barriers to consuming more fluids, and patient-selected approaches such as text messaging to maintain increased fluid intake.These component were selected based on specific behavioural change techniques (BCT)11, as noted in what follows. The “fluid prescription” was the additional daily fluid intake in excess of baseline intake required to achieve urine volume over 2.5 L/day, as recommended by American Urological Association guidelines4, and was calculated (eMethods) using the baseline 24-hour urine volume. The fluid prescription (BCT: goal setting11) was to be consumed from the smart water bottle. Participants in the intervention group were eligible for a daily, loss-framed financial incentive of $1.50 when they consumed the individualized fluid intake prescription from the smart water bottle (contingency management, BCT: material reward11). This financial incentive was available daily for the first 6 months. In months six to 18, the frequency of the days eligible for financial incentives was tapered from 80% to 15%, and no financial incentives were available in months 19 to 24 (BCT: reduce reward frequency).11, 17 Participants randomised to the intervention group were required to synchronise their water bottle and mobile device daily to be eligible for financial incentives. Participants in the intervention group were also offered structured problem solving (health coaching intervention) if the daily fluid prescription was not met for at least two 2-week periods in the first 6 months. Structured problem solving was designed to facilitate identifying barriers to increasing fluid intake and developing solutions to overcome them, prioritizing those most practicable [BCT: review behaviour goal (BCIO:007011, action planning (BCIO: 007010)].11, 17 A fidelity assessment, conducted by a trained reviewer on a sample of both initial and follow-up coaching interactions,ensured consistency of coaching across institutions.18 The fidelity assessment demonstrated that, on average, more than 90% of the required elements were covered in each initial and follow-up coaching interactions. Participants received “low touch” interventions of their choosing among social incentives (e.g., support partner) or low-cost interventions (text message reminders) to help sustain new habits of fluid consumption during months 19 to 24 (BCT: social support, prompts/cues11).

Participants in the control group were provided with guideline-concordant recommendations to increase fluid consumption to achieve urinary output of at least 2.5 L daily in addition to usual stone prevention care.4,5 They were provided with the smart water bottle, but its use was not required (eMethods). All participants in each arm continued to receive guideline-concordant care with their kidney stone clinician.

Outcomes

The primary outcome was symptomatic stone recurrence, defined as spontaneous stone passage with symptoms or a procedural intervention for a symptomatic or asymptomatic stone. Participants received questionnaires electronically to report stone events every three months for the duration of the study. Reported events were reviewed and classified as “confirmed clinical events,” “participant reported,” or “non-events” by an adjudication committee comprising expert clinicians who were blinded to study arm.19 Confirmed clinical stone events required typical stone symptoms (e.g., flank pain) in addition to objective documentation of stone passage and/or surgical intervention (Supplementary Material: Event Adjudication Charter). Events with typical symptoms and self-described passage of a stone without objective documentation of passage were classified as participant-reported; a detailed description of the adjudication process has been separately published.19 Confirmed and participant-reported events met the primary outcome. All other reported incidents were considered non-events.

Secondary outcomes included change in 24-hour urine volume (a mechanistic outcome); urinary symptoms; radiographic outcomes consisting of new stone formation and growth of an existing stone by ≥2 mm in any dimension; and a composite outcome of symptomatic stone recurrence, new stone formation, or growth of an existing stone(s). Twenty-four-hour urine collections to measure urine volume were performed at six, 12, 18, and 24 months. Participants were considered adherent if they achieved at least two 24-hour urine volumes of ≥2.5 L/day for adults or ≥30 mL/kg body weight/day for adolescents weighing <75 kg at 6, 12, 18, and 24 months. Twenty-four-hour urine collections were considered adequate if creatinine excretion rate (mg/kg/day) was within 2 standard deviations of the mean for the cohort. Urinary symptoms were recorded at baseline and every 6 months using the validated Comprehensive Assessment of Self-Reported Urinary Symptoms (CASUS).20 Radiographic outcomes were assessed using validated software that automated analyses of CT scans.21 To ensure accuracy and reliability for this trial, the performance of the analysis software was revalidated by study investigators using images from trial participants (Supplementary Material: Imaging Adjudication Charter).22 Safety monitoring included report of hyponatremia requiring hospitalization, which was evaluated systematically by review of all hospitalizations during the study.

Statistical Analysis

Statistical analysis was conducted by personnel blinded to treatment assignments. A sample size of 1642 participants (821 per group) provided at least 80% power to detect 30% relative risk reduction in stone recurrence in the intervention arm using a log-rank test for time to first symptomatic stone event, with a two-sided type I error rate of 0.05. A 15% event rate was assumed for the control group,1 and a 20% attrition was assumed over the 24-month follow-up period.

The primary outcome was evaluated according to the intent-to-treat principle and was analyzed as time-to-event by treatment group with a log-rank test. Cumulative event rates were calculated for each group as a function of time from randomisation (Kaplan-Meier method). Data collected on subsequent recurrent events of symptomatic stones were analyzed by treatment group using the Andersen-Gill model with robust standard errors to account for heterogeneity and correlation between recurrent stone events within a participant.

Twenty-four-hour urine volume measurements at baseline, 6, 12, 18, and 24 months were analyzed using a repeated-measures mixed-effects model, with study site treated as a random effect, and treatment and visit as fixed effects. Variance components were used as the variance-covariance structure. The model also included a treatment-by-visit interaction term, and denominator degrees of freedom for the F-statistic were approximated using the Kenward-Roger method.Imaging and composite outcomes were analyzed as binary outcomes at 24-months using logistic regression models. No interim analyses of the primary and secondary outcomes were planned or performed. No missing outcomes were imputed. For the primary outcome, we used a time-to-first-event approach, with participants who did not experience any stone events during the 24-month follow-up censored at the end of their observation period. For the secondary outcome of 24-hour urine volume over time (repeated measures), a mixed-effects model was used, and no imputation was required.

Pre-specified landmark analyses were conducted to exclude stone events that occurred within 30, 60, and 90 days of randomisation. These analyses were performed using the same methods as the primary outcome. Planned subgroup analyses for the primary outcome were performed using Cox proportional hazards models by including interaction terms for sex, age, clinical center, provider, and adherence. Participants were considered adherent if they achieved at least two 24-hour urine volumes of ≥2.5 L/day for adults or ≥30 mL/kg body weight/day for adolescents weighing <75 kg at six, 12, 18, and 24 months. A prespecified sensitivity analysis was performed for surgical removal of asymptomatic stones requiring ≥1 of the following criteria: stone size ≥4 mm in any dimension, mobile stone, associated hematuria, or recurrent urinary tract infection using the same methods as the primary outcome. As a sensitivity analysis, a win ratio analysis was performed using the Finkelstein-Schoenfeld method with the following hierarchical order: 1) symptomatic stone event, 2) new stone formation, and 3) stone growth. A pre-specified economic analysis will be published separately.

All statistical comparisons were performed using two-sided significance tests with α=0.05 and SAS software, version 9.4 (SAS Institute Inc., Cary, NC). All primary and secondary outcome analyses accounted for the randomization strata (age group, stone history, and study site). Full analytic plan details are provided in the statistical analysis plan (Supplementary Material: Statistical Analysis Plan).

Role of the Funding Source

This is a cooperative agreement; that means there is substantial federal scientific or programmatic involvement in the research activities. The NIDDK Project Scientist (ZK) was involved in the design and development of the clinical protocol, preparation of questionnaires and other data recording forms, coordination of research, statistical evaluations and analyses of data, and the publication of results. The program was overseen by an independent NIDDK Program Official (CM).

RESULTS

Participants

A total of 2429 patients underwent screening between September 2018 and February 2022. Of these, 1658 were randomised to intervention (n=826) or control (n=832) (Figure 1). Demographic and baseline clinical characteristics were similar between groups (Table 1). Median (25th, 75th percentiles) age was 44 (29, 59) years, and 946 (57%) participants were female. First-time stone formers represented 441/1461 (30%) of adults and 114/197 (58%) of adolescents. Baseline mean 24-hour urine volume was 1.30 (1.00–1.58) L/day in adults and 0.85 (0.66–1.10) L/day in adolescents.

Figure 1.

Figure 1.

Open in a new tab

CONSORT diagram

Table 1.

Demographic and clinical characteristics of study participants

Demographics Intervention
(n=826)		 Control
(n=832)		 Total
(n=1658)
Age, median (25th, 75th), yrs 45 (30, 60) 44 (27, 58) 44 (29, 59)
Age group, median (25th, 75th), yrs
  Adult 49 (35, 61) 48 (34, 59) 48 (36, 60)
 Adolescent 15 (14, 16) 16 (14, 16) 15 (14, 16)
Female, no. (%) 469 (56.8%) 477 (57.3%) 946 (57.1%)
Male, no. (%) 357 (43.2%) 355 (42.7%) 712 (42.9%)
Race, no. (%)
 White 732 (88.6%) 719 (86.4%) 1451 (87.5%)
 Black or African American 55 (6.7%) 58 (7.0%) 113 (6.8%)
 Native American 0 (0.0%) 3 (0.4%) 3 (0.2%)
 Asian 22 (2.7%) 25 (3.0%) 47 (2.8%)
 Native Hawaiian or Other Pacific Islander 2 (0.2%) 2 (0.2%) 4 (0.2%)
 Other/Unknowna 8 (1.0%) 11 (1.2%) 19 (1.1%)
 Multiracial 7 (0.8%) 14 (1.7%) 21 (1.3%)
Ethnicity, no. (%)
 Not Hispanic or Latino 761 (92.1%) 757 (91.0%) 1518 (91.6%)
 Hispanic or Latino 46 (5.6%) 56 (6.7%) 102 (6.2%)
 Not reported 13 (1.6%) 10 (1.2%) 23 (1.4%)
 Unknown 6 (0.7%) 9 (1.1%) 15(0.9%)
Household income, no. (%)
 Less than $90,000 323 (39.1%) 305 (36.7%) 628 (37.9%)
 $90,000 or more 365 (44.2%) 371 (44.6%) 736 (44.4%)
 Othera 138 (16.7%) 156 (18.8%) 294 (17.7%)
Thiazide diuretic (baseline), no. (%) 61 (7.4%) 60 (7.2%) 121 (7.3%)
Potassium citrate (baseline), no. (%) 92 (11.1%) 96 (11.5%) 188 (11.3%)
Baseline 24-hour urine total volume, mean (95% CI), no., L (adult) 1.29 (0.98–1.61), 727 1.30 (1.01–1.56), 734 1.30 (1.00–1.58), 1461
Baseline 24-hour urine calcium, mean (95% CI), no., mg/total volume (adult) 190 (129–261), 552 190 (127–256), 544 190 (128–259), 1096
Baseline 24-hour urine citrate, mean (95% CI), no., mg/total volume (adult) 553 (383–757),538 534 (376–729), 532 544 (378–746), 1070
Baseline 24-hour urine pH, mean (95% CI), no., (adult) 6.00 (5.65–6.41), 671 6.04 (5.66–6.41), 685 6.02 (5.65–6.41), 1356
Baseline 24-hour urine osmolality, mean (95% CI), no., mOsm/kg 662 (471–847), 432 653 (463–815), 469 657 (469–834), 901
Baseline 24-hour urine sodium, mean (95% CI), no., mEq/TV (adult) 140 (103–183), 549 134 (99–177), 540 137 (101–179),1089
Baseline 24-hour urine potassium, mean (95% CI), no., mEq/TV (adult) 50 (36–66), 538 48 (36–62), 525 49 (36–64), 1063
Open in a new tab
a

Includes ‘don’t know’ and ‘prefer not to answer’ responses.

Primary Outcome

At a median follow-up of 738 (711, 778) days, 154 (18.6%) participants experienced a symptomatic stone event in the intervention group compared with 165 (19.8%) in the control group (hazard ratio [HR] 0.96; 95% confidence interval [CI] 0.77 to 1.20). There was no difference in the cumulative risk of symptomatic stone recurrence between the groups (Figure 2). Subgroup analyses showed no heterogeneity of treatment effect by sex, age, clinical center, or provider. In addition, adherence to the intervention, defined as meeting at least two 24-hour urine collections at or above target (2.5L/day for adults), did not result in a difference in the adherence subgroups (Figure 3, Supplementary Appendix: Tables S1 & S2). In sensitivity analyses, there was no difference between groups when excluding those who underwent surgery for an asymptomatic stone or stone events that occurred within 30, 60, and 90 days of randomisation (Supplementary Appendix: Table S3). A hierarchical order (win ratio) analysis did not show a difference between treatment groups (Supplementary Appendix: Table S4).

Figure 2.

Figure 2.

Open in a new tab

Symptomatic recurrence of urinary stones

Figure 3.

Figure 3.

Open in a new tab

Subgroup analyses of symptomatic recurrence of urinary stones

*Adherence defined as at least 2.5 L UOP/24 hours on at least two 24-hour collections for adult participants. Additionally, the 24-hour urine collection is required to have creatinine mg/kg/day within 2 SD of mean Cr mg/kg/day for the cohort.

Secondary Outcomes

Twenty-four-hour urine volume increased from baseline in both groups and was higher in the intervention group at months six, 12, 18, and 24 compared with the control group (Figure 4, Supplementary Appendix: Table S5). Urinary storage symptoms of frequency, urgency, and nocturia were greater in the intervention group versus control at months six (p=0.050) and 12 (p=0.014) but not at other time points.

Figure 4.

Figure 4.

Open in a new tab

24-hour urine volume by time point and treatment group (mixed-effects model)

Fin Incent (%)=Loss-framed financial incentive. % indicates percentage of days loss-framed incentive was available to participants.

FI Taper=Financial incentive taper. Availability tapered from 75% of days (month 13) to 15% of days (month 18). Participants blinded to which day(s) financial incentive was available. No financial incentive available months 19–24.

SPS=Structured problem solving.

Low Touch=“Low touch” interventions to promote adherence (e.g., support partner, reminder via text communication).

There was no difference in stone growth ≥2 mm (OR 1.44; 95% CI 0.91 to 2.29) or new stones (OR 0.99; 95% CI 0.72 to 1.37) between groups from baseline to end-of-study imaging (Supplementary Appendix: Table S6). The composite outcome of symptomatic stone recurrence, new stone formation, or stone growth was also not statistically different between treatment groups.

Safety

No participants experienced hyponatremia requiring hospitalization. In the intervention group, 12 (1.5%) participants had asymptomatic hyponatremia versus two (0.2%) participants in the control group (p=0.018). There were no study-related serious adverse events in either group.

DISCUSSION

The PUSH trial is the largest randomised controlled trial designed to improve adherence to fluid intake for secondary prevention of urinary stones to date. There are several key findings. Participants in the behavioural intervention arm achieved a greater urine volume compared with participants in the control arm. However, this increase in urine volume did not result in a decrease in recurrent symptomatic stone events over two years of follow-up. Second, urinary symptoms were greater in the intervention group at six and 12 months. Finally, there was no difference between groups in the radiographic outcomes of new stone formation or stone growth at end of study.

Increasing fluid intake to increase urine volume has long been a cornerstone of secondary prevention of urinary stone disease, with the goal of diluting the concentration of stone-forming salts in urine. Both the American Urological Association4 and the European Association of Urology guidelines5 recommend fluid intake sufficient to produce a urine volume of ≥2.5 L/day as a prevention strategy for patients with urinary stone disease. These recommendations are partially based on a single randomised controlled trial of 199 participants which demonstrated that greater fluid intake resulting in 2.5L of urine/day on average led to fewer recurrent stones and longer time to stone events compared with a group that made an average of only 1L of urine per day.7 Largely based on the reduction in recurrent stone events reported in that trial, increased fluid intake is now considered standard for the secondary prevention of urinary stone disease.4,5 However, in real-world care, it is challenging for patients to increase their urine volume to meet this goal, despite the efforts of various providers in multiple clinical settings.8

In the PUSH trial, adult participants in the intervention arm increased urine volume (versus baseline) by 600 mL/24 hours at 6 months, compared with only 360 mL/24 hr in the control arm, but this difference diminished over time. An important aspect of the PUSH trial was that participants in the control arm continued to receive ongoing care from their existing stone clinician which included increasing fluid intake and dietary recommendations per their usual practice, compared to the Borghi trial7 in which the control arm was randomised to no additional fluids. Accordingly, the increase in urine volume in the control group in our study and the resulting smaller difference in urine volume between the two groups (Figure 4) may explain the similarity between groups for stone-related outcomes.

There are myriad perceived barriers to fluid intake such as not liking the taste of water, not experiencing thirst, failure of habit formation for fluid intake, feeling bloated with consumption, while many do not understand the association between fluid intake and stone formation.9 In this study, we attempted to address barriers to fluid intake through health coaching and structured problem solving, theorizing that patient education and providing individualized solutions would optimize success of the intervention. However, the difference in urine volume between groups remained modest. This outcome may, in part, reflect a unique negative effect of the intervention for this trial in that urinary symptoms are often exacerbated by increased fluid intake. Participants assigned to the intervention reported greater urinary symptoms at months 6 and 12, time points at which the difference in urine volume between the two arms was greater (Supplementary Appendix 2: Table S7). As fluid intake increases, urine output is expected to rise, which in some patients may result in expected, but potentially bothersome, urinary symptoms such as frequency or nocturia. These parallel increases in urine volume and urinary symptoms may reflect the challenges that affect some stone formers when trying to adhere to an increased fluid intake goal.

Several randomised controlled trials have demonstrated that interventions can improve healthy behaviours such as smoking cessation, weight loss, and physical activity.12.24,25 Broad adoption of these approaches has been hindered by two key limitations: durability of the health behaviour change and the use of surrogate rather than clinically meaningful outcomes. We designed the PUSH trial to address these limitations.17 First, we used a clinically meaningful primary outcome (symptomatic stone recurrence) rather than the mechanistic surrogate outcome of increased urine volume.25 Second, we structured the intervention to enhance durability of the effect through habit formation using behavioural approaches known to improve adherence, including an initial loss-framed incentive, tapering of incentives with intermittent reinforcement,11,26 and additional components that operate on structural barriers not addressed by financial incentives. Adaptable patient-selected strategies allowed for individualization of plans to maximize success of the intervention. The positive effect on a surrogate, but less patient-relevant outcome (24-hour urine volume) highlights the importance of following interventions to meaningful clinical endpoints in adherence trials and should inform the design of future health behaviour trials.

This study has limitations. Participants were recruited from academic, tertiary care centers that may treat participants with more severe disease. We did not assess or control for additional fluid intake not measured by the smart water bottle or other dietary factors, although randomisation should balance these unmeasured confounders. The follow-up period was only two years; however, the PUSH trial exceeded the estimated event rate, and the Kaplan-Meier curves were nearly parallel, suggesting that a longer follow-up period for PUSH would be unlikely to change the results. Insight into which components of the intervention were effective (or not) would have required a larger study design (e.g., factorial), which was not feasible. These factors could be assessed in an exploratory secondary analysis. Female participants made up 57% of the study population. Although the prevalence of urinary stone disease among females in the United States is closer to 44%, the gender gap is narrowing worldwide.27 The predominance of female participants may reflect gender differences in willingness to participate in clinical research. In particular, females are more likely than males to enroll in behavioural intervention trials28 (median 56.7%; 95% CI 40.7%,76.0% [n=1346 trials]). Utilization of adjunctive stone prevention medications (thiazides, potassium citrate) was similar between groups at baseline (Table 1). We did not assess participant acceptability with the intervention, though the dropout rate was lower than projected. Finally, there is the risk of co-intervention through treatment in a specialty stone clinic, known as the “stone clinic effect.”29 This effect, along with frequent study contact with the control group participants, could have biased results toward the null. These limitations notwithstanding, several features of the trial strengthen the validity of outcomes. The study population was the largest to date for a stone prevention trial, with adequate power to detect meaningful differences, and the number of primary outcome events exceeded projections. Additionally, we exceeded the target enrollment, with an attrition rate lower than projected. Multiple complementary methods were used to ascertain outcomes to reduce underreporting of events.

Although there was no difference in stone recurrence between the groups, this trial has important implications for behavioural science and for stone prevention efforts. This intervention increased urine volume but did not decrease the long-term clinical goal of stone recurrence. The results of the PUSH study do not undermine the importance of increasing fluid intake for stone prevention, as this remains a low-cost, low-risk intervention with likely benefits based on prior literature. Rather, our study suggests that while contingency management intervention with health coaching for non-adherent participants was insufficient to yield significant clinical benefit, relative to control participants, future investigations could apply optimization study designs [i.e., factorial designs, or sequential, multiple assignment, randomized trial (SMART) designs] within a multi-phase optimization strategy (MOST) framework to determine which strategies yield the best results for most people. Additional qualitative research may yield key insights into patient barriers to fluid intake that would inform future intervention design. Potentially, a more personalized medicine approach could be attempted in order to address individual differences in motivations to consume water and thereby prevent recurrent stones.

Supplementary Material

Supplementary Material

Research in context.

Evidence before this study

We performed a systematic search of PubMed for randomized trials testing interventions to promote adherence to fluid intake for secondary prevention of urinary stone disease (USD). We searched PubMed (from Jan 1, 1995 to June 2, 2025) and the detailed search strategy is available online: https://osf.io/asf3z/. We also searched clinicaltrials.gov (through 6/2/25) to identify similar trials in progress or without published results. We identified 222 publications and 5 clinicaltrials.gov listings. Of these, we identified a total of 6 trials (including one published only as a conference abstract) testing adherence interventions for fluid intake for secondary stone prevention.

Of these 6 trials, two were initiated prior to 2017, when our Prevention of Urinary Stones with Hydration (PUSH) began. The Hidrate Me study (NCT02938884) randomized participants with a history of USD and low urine volume to a smart water bottle (Hidrate Spark) versus standard water bottle, facilitating self-monitoring of behaviour in the intervention arm. The results of this trial were reported in 2022 (PMID 35283036). Among 85 participants enrolled, 51 (60%) participants completed 24-hour urine collections at 6 weeks, with a greater increase in urine volume over baseline in the smart water bottle arm compared to control (+1.37 L/day vs +0.79 L/day, p=0.04). Stone recurrence was not ascertained as an outcome.

The second study initiated prior to 2017 was an observational cohort study (NCT01928108) enrolling adults with a history of USD which compared use of two different smart phone applications to manually track water consumption, again employing a behaviour change technique of self-monitoring. Neither the literature search nor clinicaltrials.gov reported study results.

The other 4 trials of adherence interventions for fluid intake in USD prevention were initiated after PUSH started. The protocol describing the sipIT2 trial (with a primary endpoint of 24-hour urine volume) was published in 2024 (PMID: 38253254). Two RCTs identified only via clinical trials.gov compared a smartphone application delivering prevention education versus control arm receiving no stone prevention information. The sixth trial (comparing mobile app care plan versus “standard kidney stone follow-up pathway”) was identified only via a conference abstract; no clinicaltrials.gov registration was identified.

Current guidance for water intake for secondary stone prevention is based on a Cochrane Systematic Review (2020) that found a single RCT (Borghi, 1996) comparing the effects of high water intake vs low water intake for secondary USD prevention. Cochrane assessed this RCT as low certainty evidence.

Added value of this study

The PUSH trial advances the field by testing an adaptive, multi-component behavioural health approach to promoting fluid intake adherence in a large patient population of patients with USD and low urine volume. The PUSH intervention is grounded in contingency management, leveraging a loss-framed financial incentive, along with an adaptive structured problem -solving intervention that is responsive to participant non-adherence. In addition to examining 24-hour urine output, PUSH is the first adherence study to assess the clinical endpoint of stone recurrence.

Implication of all available evidence

Very limited evidence is available to guide clinicians about the best strategies that will increase and sustain high fluid intake for secondary USD prevention. The PUSH trial results suggest that for many patients with USD and low urine volume, it may be difficult to sustain high fluid intake and thereby reduce stone recurrence. Taken together, these results suggest that investigators may need to focus on alternative adherence strategies and/or secondary prevention strategies that go beyond simply increasing fluid intake.

Acknowledgments

Elizabeth E.S. Cook of the Duke Clinical Research Institute provided editorial assistance. Sarah Cantrell of the Duke Medical Center Library provided assistance with the literature search.

Urinary Stone Disease Research Network: The following individuals were instrumental in the planning and conduct of the PUSH trial at each of the participating institutions:

Clinical Centers

University of Pennsylvania/Children’s Hospital of Pennsylvania, Philadelphia, PA: PIs: Peter P. Reese, MD, PhD, Gregory E. Tasian, MD MSCE; Co-Is: Sandra Amaral, MD, MHS, Janet Audrain-McGovern, PhD, Justin Ziemba, MD, Kevin Volpp, MD; Project Manager: Adam Mussell; Study Coordinators: Emily Funsten, Gabrielle Perez, Brittney Henderson, Kristen Koepsell.

University of Texas Southwestern Medical Center, Dallas, TX: PI: Naim M. Maalouf, MD; Co-Is: Jodi A. Antonelli, MD, Linda A. Baker, MD, Lakshmi Ananthakrishnan, MD; Local Referring Providers: Brett A. Johnson, MD, Yair Lotan, MD, Orson W. Moe, MD, Margaret S. Pearle, MD, PhD, Craig A. Peters, MD, Khashayar Sakhaee, MD, Li Song, MD; Referring Collaborators (for participants recruited remotely): Timothy Y. Tseng, MD (University of Texas Health-San Antonio), Ryan L. Steinberg (University of Iowa), Joseph J. Crivelli (University of Alabama); UT Southwestern Study Staff: Sudeepa Bhattacharya (Program manager), Martinez Hill (CRC), Esperanza Jackson (CRC), Alejandra Lozano (CRC), Corey Nixon (CRC), Joyce Obiaro (CRC), Brooke Piskator (CRC), Cynthia Rangel (CRC), Jesse Tarbutton (CRC), Madeline Worsham (Coach), John R. Poindexter (Local Database Manager). UT Southwestern CTSA Program, grant UL1TR003163

University of Washington, Seattle, WA: PIs: Jonathan D. Harper, MD, Hunter Wessells, MD; Co-Is: Fionnuala Cormack, MD, Mathew Sorensen, MD, Karyn Yonekawa, MD; Study Coordinators: Holly Covert, Tristan Baxter, Elsa Ayala.

Washington University in St. Louis, St. Louis, MO: PIs: Alana C. Desai, MD, H. Henry Lai, MD; Co-Is: Vincent Mellnick, MD, Douglas Coplen, MD; Study Coordinators: Juanita Taylor, Aleksandra Klim, Deborah Ksiazek, Vivien Gardner; Referring Surgeon: Niraji Bhadiwala, MD.

Recruiting Centers

Cleveland Clinic Foundation, Cleveland, OH: PI: Sri Sivalingam, MD, MSc, FRCSC; Co-Is: Katherine Dell, MD, Juan Calle, MD, Manoj Monga, MD, Louisa Ho, MD, Harmenjit Brar, MD; Referring Practitioners: Heidi Digennaro, CNP, Tiffany Loboda, CNP; Study Coordinators: Paige Gotwald, Marina Markovic.

Mayo Clinic Foundation, Rochester, MN: PI: John Lieske, MD; Co-Is: Kevin Koo, MD, Fernanda Bellolio, MD, Michelle Bouquet, PA-C, Andrew Rule, MD, Stephen Erickson, MD, Mira Keddis, MD, Aaron Potrezke, MD, Andrea Ferrero, PhD, David Sas, DO; Study Coordinators: Angela Waits, Courtney Lenort.

Scientific Data Research Center

Duke Clinical Research Institute, Duke University, Durham, NC: PIs: Charles D. Scales, Jr, MD, MSHS, Hussein R. Al-Khalidi, PhD; Co-Is: Kevin Weinfurt, PhD, Hayden Bosworth, MD; Statistician: Hongqiu Yang, PhD; Project Leadership: Laura Johnson, Davy Andersen, Paul Camarena; Lead CRA: Sharon Settles; CRA: Angela Venetta; Data Manager: Omar Thompson, Robert Baldwin.

National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)

Project Scientist: Ziya Kirkali, MD; Program Official: Christopher Mullins, PhD.

Data and Safety Monitoring Board: John Denstedt, MD (DSMB Chair), Uri Alon, MD, Dean G. Assimos, MD, Scott Cohen, MD, Gary Curhan, MD, Michael A. Freeman, MD, David Goldfarb, MD, Amy Krambeck, MD, Rebecca A. Krukowski, PhD, Jeannette Lee, PhD, Manoj Monga, MD, Andrew Rule, MD, Christopher Schmid, PhD, Marshall Stoller, MD, Eric Taylor, MD, Jennifer Temple, PhD.

Declaration of Interests

Desai: No disclosures

Maalouf: No disclosures

Harper: No disclosures

Sivalingam: No disclosures

Lieske: No disclosures

Lai: No disclosures

Reese: No disclosures

Wessells: No disclosures

Yang: No disclosures

Al-Khalidi: No disclosures

Kirkali: No disclosures

Tasian: No disclosures

Scales: DSMB Chair, Stent Omission after Ureteroscopic Lithotripsy trial, University of Michigan, funding from PCORI

Funding

This research was funded by the National Institute of Diabetes and Digestive and Kidney Diseases DK110986 (WUSTL); DK110961 (CHOP/UPENN); DK110954 (UW); DK110994 (UTSW); DK110988 (Duke).

Role of Funding Source

The research was funded by the National Institute of Diabetes and Digestive and Kidney Diseases. This is a cooperative agreement; that means there is substantial federal scientific or programmatic involvement in the research activities. The NIDDK Project Scientist (ZK) is involved in the design and development of the clinical protocol, preparation of questionnaires and other data recording forms, coordination of research, statistical evaluations and analyses of data, and the publication of results. The program is overseen by an independent NIDDK Program Official.

Data Sharing Statement

After publication, de-identified data from the trial database will be transferred to the NIDDK Central Data & Biorepository (https://repository.niddk.nih.gov/home). Proposals for data access can be submitted via email (USDRN@duke.edu) or directly to the NIDDK Central Data & Biorepository. Data will be made available to those with an approved proposal and executed data access agreement.

References

  • 1.Rule AD, Lieske JC, Li X, Melton LJ III, Krambeck AE, Bergstralh EJ. The ROKS nomogram for predicting a second symptomatic stone episode. J Am Soc Nephrol. 2014;25(12):2878–2886. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Bryant M, Angell J, Tu H, Goodman M, Pattaras J, Ogan K. Health Related Quality of Life for Stone Formers. J Urol. 2012;188(2):436–440. [DOI] [PubMed] [Google Scholar]
  • 3.Moe OW. Kidney stones: pathophysiology and medical management. Lancet. 2006. Jan 28;367(9507):333–44. doi: 10.1016/S0140-6736(06)68071-9. [DOI] [PubMed] [Google Scholar]
  • 4.Pearle MS, Goldfarb DS, Assimos DG, et al. Medical management of kidney stones: AUA Guideline. J Urol. 2014;192(2):316–324. [DOI] [PubMed] [Google Scholar]
  • 5.Skolarikos A, Jung H, Neisius A, et al. EAU Guidelines on Urolithiasis. EAU Guidelines. Edn. presented at the EAU Annual Congress Paris 2024. ISBN 978-94-92671-23-3.
  • 6.Bao Y, Tu X, Wei Q. Water for preventing urinary stones. Cochrane database of systematic reviews, 2020, Issue 2. Art no. CD004292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Borghi L, Meschi T, Amato F, Briganti A, Novarini A, GianniniA. Urinary volume, water, and recurrences in idiopathic calcium nephrolithiasis: a 5-year randomised prospective study. J Urol. 1996;155(3):839–843. [PubMed] [Google Scholar]
  • 8.Parks JH, Goldfischer ER, Coe FL. Changes in urine volume accomplished by physicians treating nephrolithiasis. J Urol. 2003;169(3):863–866. [DOI] [PubMed] [Google Scholar]
  • 9.Rice P, Archer M, Davis T, Pietropaolo A, Somani B. Patient perception and barriers with fluid hydration: a prospective face-to-face interview and counselling from a university hospital stone clinic. Cent European J Urol. 2023; 76(3)239–244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.US Preventive Services Task Force. Behavioral Weight Loss Interventions to Prevent Obesity-Related Morbidity and Mortality in Adults: US Preventive Services Task Force Recommendation Statement. JAMA 2018;320(11):1163–1171. [DOI] [PubMed] [Google Scholar]
  • 11.Michie S, Richardson M, Johnston M, et al. The behaviour change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behaviour change interventions. Ann Behav Med. 2013. Aug; 46(1):81–95. [DOI] [PubMed] [Google Scholar]
  • 12.Volpp KG, John LK, Troxel AB, Norton L, Fassbender J, Loewenstein G. Financial incentive–based approaches for weight loss: A randomized trial. JAMA. 2008;300(22):2631–2637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Secades-Villa R, Gonzalez-Roz A, Vallejo-Seco G, et al. Additive effectiveness of contingency management on cognitive behavioural treatment for smokers with depression:Six-month abstinence and depression outcomes. Drug and Alcohol Dependence 2019; 204: 107495. [DOI] [PubMed] [Google Scholar]
  • 14.Almirall D, Nahum-Shani I, Sherwood N, et al. Introduction to SMART designs for the development of adaptive interventions: with application to weight loss research. Transl Behav Med. 2014;4(3):260–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Yan X, Dunne DM, Impey SG et al. A pilot sequential multiple assignment randomized trial (SMART) protocol for developing an adaptive coaching intervention around a mobile application for athletes to improve carbohydrate periodization behavior. Con Clin Trial Comm 2022; 26:100899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Ciani O, Buyse M, Garside R, et al. Comparison of treatment effect sizes associated with surrogate and final patient relevant outcomes in randomised controlled trials: meta-epidemiological study. BMJ. 2013;346:f457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Scales CD, Desai AC, Harper JD, et al. Prevention of Urinary Stones with Hydration (PUSH): Design and Rationale of a Clinical Trial. Am J Kidney Dis. 2021;77(6):898–906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Reese PP, Shah S, Funsten E, et al. Using structured problem solving to promote fluid consumption in the prevention of urinary stones with hydration (PUSH) trial. BMC Nephrol. 2024;25(1):183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Wessells H, Lieske JC, Lai HH, et al. Adjudication of Self-reported Symptomatic Stone Recurrence in the Prevention of Urinary Stones with Hydration Trial. Urology. 2024;194:27–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Weinfurt KP, Griffith JW, Flynn KE, et al. The Comprehensive Assessment of Self-Reported Urinary Symptoms: A New Tool for Research on Subtypes of Patients with Lower Urinary Tract Symptoms. J Urol. 2019;201(6):1177–1183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Babajide R, Lembrikova K, Ziemba J, Ding J, et al. Automated Machine Learning Segmentation and Measurement of Urinary Stones on CT Scan. Urology. 2022;169:41–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Tasian GE, Maalouf NM, Harper JD, et al. Validation of an Automated CT Image Analysis in the Prevention of Urinary Stones with Hydration Trial. J Endourol. 2025;39(9):953–959.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Charness G, Gneezy U. Incentives to exercise. Econometrica. 2009;77(3):909–931. [Google Scholar]
  • 24.Volpp KG, Troxel AB, Pauly MV, et al. A randomized, controlled trial of financial incentives for smoking cessation. N Eng J Med. 2009;360(7):699–709. [DOI] [PubMed] [Google Scholar]
  • 25.Stout RE, Lingeman JE, Krambeck AE, et al. A randomized trial evaluating the use of a smart water bottle to increase fluid intake in stone formers. Journal of Renal Nutrition 2022;32(4):389–395. [DOI] [PubMed] [Google Scholar]
  • 26.DeRusso AL, Fan D, Gupta J, Shelest O, Costa RM, Yin HH. Instrumental uncertainty as a determinant of behavior under interval schedules of reinforcement. Front Integr Neurosci. 2010;4:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Gillams K, Juliebø-Jones P, Juliebø SØ, Somani BK. Gender Differences in Kidney Stone Disease (KSD): Findings from a Systematic Review. Curr Urol Rep. 2021;22(10):50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Steinberg JR, Turner BE, Weeks BT, Magnani CJ, et al. Analysis of Female Enrollment and Participant Sex by Burden of Disease in US Clinical Trials Between 2000 and 2020. JAMA Netw Open. 2021;4(6):e2113749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hosking DH, Erickson SB, Van Den Berg CJ, Wilson DM, Smith LH. The Stone Clinic Effect in Patients with Idiopathic Calcium Urolithiasis. J Urol. 1983;130(6):1115–1118. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS2159793-supplement-Supplementary_Material.pdf (3.3MB, pdf)

Data Availability Statement

After publication, de-identified data from the trial database will be transferred to the NIDDK Central Data & Biorepository (https://repository.niddk.nih.gov/home). Proposals for data access can be submitted via email (USDRN@duke.edu) or directly to the NIDDK Central Data & Biorepository. Data will be made available to those with an approved proposal and executed data access agreement.

RESOURCES