Abstract
Background and Objective: Urea is an alternative for treatment of hyponatremia however, its use has not been widely studied. The purpose of this study was to evaluate the safety and efficacy of urea for the treatment of hyponatremia. Methods: A retrospective cohort of patients with hyponatremia (serum sodium <135 mEq/L) of any cause who received at least 1 dose of urea during hospitalization and no prior use of urea. Serum sodium levels were collected at baseline and for 4 days or until urea was discontinued, whichever occurred first. The primary outcome was the serum sodium change between baseline and discharge or urea discontinuation. Results: Median serum sodium increased 2 [IQR, 0-4] mEq/L per day after urea administration at a median dose of 30 g/day. A significant difference in serum sodium was observed between baseline and discharge or discontinuation (124.2 ± 4 vs 130.1 ± 5.1; P < .001) and serum blood urea nitrogen (BUN) levels (18.4 ± 13.1 vs 41.1 ± 26.6; P ≤ .001). Serum sodium overcorrection (increase >8 mEq/L in 24 hours) occurred in 6 patients (8%). Urea was discontinued in 39 patients (53%); 20 discontinuations were due to patient intolerance. Conclusion: Urea appears to be an effective treatment for hyponatremia; however, patient tolerance, the rate of serum sodium overcorrection, and outpatient affordability may limit its use.
Keywords: cost effectiveness, drug/medical use evaluation, fluid and electrolyte disorders, formulary management/P & T, genitourinary, monitoring drug therapy, nephrology
Background
Hyponatremia is defined as serum sodium <135 mEq/L and is the most common disorder of fluid and electrolyte balance. It is associated with high morbidity and mortality and occurs in 15% to 30% of hospitalized patients. Common causes of hyponatremia include syndrome of inappropriate antidiuretic hormone (SIADH) and fluid imbalances related to cardiovascular or renal conditions. It is further categorized based on urine osmolality (hypertonic or hypotonic) and patient volume status (hypovolemic, euvolemic, or hypervolemic). Usual first-line treatment options depend on the etiology of hyponatremia and include fluid restriction, loop diuretics, sodium chloride tablets or infusions, withholding offending agents, and vasopressin antagonists (tolvaptan and conivaptan). 1 The use of oral urea for hyponatremia was first reported in 1981. 2 It was historically administered intravenously to reduce intracranial pressure but fell out of favor due to rebound intracranial hypertension due to its ability to cross the blood brain barrier (BBB). 3
There is renewed interest in oral urea after Decaux and Genette 4 reported positive outcomes for use in long-term treatment of hyponatremia associated with SIADH. Urea is a small, soluble, and diffusible molecule that is filtered in the glomerulus and excreted in the urine along with electrolyte-free water. 5 It is classified as a medical food by the Food and Drug Administration (FDA) and is commercially available in lemon-lime flavored 15 g powder packets. The usual initial dose ranges from 15 to 60 g per day divided twice daily, and it is recommended to carefully monitor serum sodium while on therapy. Side effects include nausea, vomiting, diarrhea, dysgeusia, and volume depletion.5,6 Serum sodium overcorrection has been reported with urea therapy, defined as an increase in serum sodium >10 mmol/L in 24 hours or >8 mmol/L in 24 hours for patients at high risk for osmotic demyelination syndrome (ODS), including advanced liver disease, serum sodium <105 mmol/L, alcoholism, malnutrition, and hypokalemia. 1 It is important to note overcorrection of hyponatremia with urea has shown significantly lower mortality and central nervous system impairment such as ODS than overcorrection caused by vasopressin antagonists and hypertonic saline; however, this has only been observed in animal models. 7 The lower rate of overcorrection is thought to be due to reduced BBB permeability with urea compared to vasopressin antagonists and hypertonic saline, but this has not been well defined. BUN levels may increase following urea administration, but this isolated increase is not indicative of kidney dysfunction. 5 Urea was added to the formulary at the study site for treatment of hyponatremia in April 2018. Though it is classified as a medical food and not a medication, the Pharmacy and Therapeutics Committee recommended this product be procured and dispensed by the pharmacy department due to clinical impact on serum sodium.
The 2013 United States (U.S.) expert panel and the 2014 European clinical guidelines recommend urea as a second line alternative behind vasopressin antagonists for chronic hyponatremia due to SIADH.1,8 Urea has not been studied in randomized controlled trials; however, it has demonstrated potential benefit in several small studies. Two single-center retrospective cohort studies have evaluated urea’s use in critically ill patients with hyponatremia.4,9 Urea has also been evaluated for long-term use in a year-long crossover trial. 10 Lastly, urea demonstrated efficacy in the inpatient setting in non-critically ill adults in 2 recent retrospective studies.11,12 To our knowledge, only 1 previous study has evaluated urea for use in short-term treatment of hyponatremia that is not secondary to SIADH. 11
Although data on urea is less robust than vasopressin antagonists, it may be more cost-effective given the rising cost of prescription drugs in the U.S. The typical average wholesale price (AWP) for urea is $7.42 per day (15 g twice daily) compared to $495.43 to $513.95 per day for tolvaptan and $656.33 to $1312.66 for conivaptan standard maintenance doses.6,13,14 It is important to note that urea may not be accepted by outpatient insurance plans as a medical food which may translate to increased out of pocket costs. While the results of the existing studies are promising, data to support the use of urea for the treatment of hyponatremia are still lacking. The objectives of this study were to evaluate the safety and efficacy of urea in the treatment of hyponatremia.
Methods
This single-center, retrospective study was conducted evaluating adult patients at a large academic medical center between October 2018 and November 2019. The study included patients who received at least 1 dose of urea while hospitalized and excluded those who were taking urea prior to hospitalization as indicated by medication history.
Data were extracted retrospectively from the electronic medical record, EPIC Systems Corporation® (Verona, WI), and were collected and managed using the Research Electronic Data Capture (REDCap) electronic data capture tools hosted at the study site.15,16 REDCap is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for data integration and interoperability with external sources. This publication was supported by Award Number Grant UL1TR002733 from the National Center for Advancing Translational Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Advancing Translational Sciences or the National Institutes of Health.15,16 Institutional Data Quality Release approval was granted by the Chief Quality Officer.
Urea is ordered in the electronic health record through a standardized order panel defaulting to urea 15 g twice daily along with serum sodium collection every 12 hours. Of note, clinicians may opt for different urea dosing and/or more frequent serum sodium monitoring. Serum sodium (mEq/L) was extracted at baseline, then every 12 hours for 2 days, then daily for a total of 4 days or until day of discharge or urea discontinuation, whichever occurred first. BUN (mg/dL) and urea dose (g/day) were extracted for a total of 4 days or until day of discharge or urea discontinuation, whichever occurred first. Baseline and daily values were those collected with morning labs on the day of urea initiation. Discharge values were the final values collected prior to discharge. If urea was discontinued prior to discharge, the final serum sodium prior to discharge was extracted.
Laboratory values were excluded if they were collected less than or greater than 4 hours from intended collection time. If laboratory values were obtained more frequently than collected in our study (eg, every 8 hours), we collected those obtained with morning labs and 12 hours thereafter for serum sodium. Etiology of hyponatremia and indication for urea treatment were assessed by review of progress notes. Concomitant treatment for hyponatremia was included if it was administered within 5 days prior to urea initiation; however, the duration of concomitant therapy was not collected. The primary efficacy outcome was the serum sodium change between baseline and discharge or urea discontinuation. A secondary efficacy outcome was the difference in serum sodium at 24 hour intervals after administration (eg, 24-48 hours) through day 4. The primary safety outcome was the occurrence of overcorrection of serum sodium, defined as an increase of >8 mEq/L in 24 hours. Incidence of hypernatremia, defined as a serum sodium >145 mEq/L, was also collected. For purposes of our study, serum sodium is reported as mEq/L which is equivalent to mmol/L reported in previous studies and guidelines. A secondary safety outcome was the BUN change between baseline and discharge or discontinuation after urea administration. Discontinuation of urea along with reason(s) for discontinuation and adverse effects were collected from review of progress notes. Continuous data were analyzed using the paired Student t-test for parametric data and Wilcoxon signed rank test for nonparametric data. Categorical data were analyzed using either the χ2 test or Fisher exact test. Normally distributed data are presented as mean (± standard deviation) and non-normally distributed data are presented as median (25%-75% interquartile range [IQR]). All tests were 2-tailed, and P < .05 was used to represent statistical significance. A sub-analysis of the primary aim was performed using the Student t-test to assess differences in those who received urea alone or with concomitant therapy. Statistical analysis was performed using IBM SPSS Statistics (version 26.0, Armonk, NY, IBM Corp).
Results
A total of 74 patients were included in this retrospective analysis out of 77 eligible patients between October 2018 and November 2019. Three patients were excluded; 2 were on urea prior to admission and 1 never received a documented dose. The median age was 67 years [IQR, 60-80], with 46% being female. The most common etiology for hyponatremia was SIADH, affecting 44 (59%) patients. Thirty-nine (53%) patients received urea alone. For patients who received concomitant treatment, normal saline was the most common therapy, administered to 19 patients (54%). Table 1 displays demographic, etiology, and concomitant therapy data. Urea dosing ranged from 7.5 to 90 g/day and median urea dosing was 30 g/day [IQR, 15-45] (Table 2).
Table 1.
Baseline Characteristics (n = 74).
| Variable | Value |
|---|---|
| Demographics | |
| Age, years | 67 [60-80] |
| Length of stay, days | 15 (± 15) |
| Female, n (%) | 34 (46) |
| Etiology of hyponatremia | |
| Euvolemic, n (%) | 49 (66) |
| SIADH | 44 |
| Unable to identify | 5 |
| Hypervolemic, n (%) | 6 (8) |
| Heart failure | 2 |
| Unable to identify | 2 |
| Cirrhosis | 1 |
| Chronic kidney disease | 1 |
| Hypovolemic, n (%) | 4 (6) |
| Unable to identify | 3 |
| Renal losses | 1 |
| Unknown, n (%) | 15 (20) |
| Hyponatremia treatment | |
| Urea alone, n (%) | 39 (53) |
| Concomitant therapy, n (%) | 35 (47) |
| 0.9% sodium chloride infusion | 19 |
| Sodium chloride tablets | 12 |
| 3% sodium chloride infusion | 5 |
| Loop diuretics | 5 |
| Vasopressin antagonists | 2 |
| Held outpatient diuretics | 2 |
Note. Nominal data are presented as number (percentage) where applicable. Continuous data are presented as mean ± standard deviation (SD) if parametric or median [interquartile range (IQR)] if non-parametric. SIADH = syndrome of inappropriate antidiuretic hormone
Table 2.
Urea Dosing.
| Day of therapy | Min | Max | Median [IQR] |
|---|---|---|---|
| Day 1 | 7.5 | 60 | 30 [15-30] |
| Day 2 | 15 | 90 | 30 [15-45] |
| Day 3 | 15 | 90 | 30 [15-45] |
| Day 4 | 15 | 90 | 30 [30-45] |
| Discontinuation/discharge | 15 | 90 | 30 [30-45] |
Note. Data presented as g/day. Continuous data are non-parametric and are presented as median [interquartile range (IQR)].
Median serum sodium increased 2 [IQR, 0-4] mEq/L per day after urea administration. A significant difference in serum sodium was observed between baseline and discharge or discontinuation serum sodium in 74 patients (124.2 ± 4 vs 130.1 ± 5.1; P < .001). This finding may not be clinically significant due to the overlap in standard deviation between the 2 groups. Similar results were found in changes of serum sodium between baseline and 12 hours in 63 patients (124.1 ± 4.2 vs 125.2 ± 4.6; P = .01), baseline and 24 hours in 65 patients (124.2 ± 4.0 vs 126.5 ± 4.7; P < .001), 24 to 48 hours in 46 patients (125.7 ± 4.5 vs 128.2 ± 4.3; P < .001), 48 to 72 hours in 30 patients (128.3 ± 4.5 vs 130.6 ± 4.0; P < .001), and 72 to 96 hours in 17 patients (130.4 ± 3.8 vs 132.4 ± 3.7; P = .003). We observed missing data points at 12 hours (n = 11) and 24 hours (n = 9) due to urea discontinuation, lab timing, or lab omission.
A total of 6 patients (8%) experienced serum sodium overcorrection within the initial 48 hours of urea therapy (Table 3). There were no occurrences of hypernatremia or ODS noted on retrospective chart review. Serum BUN levels increased from a median of 16 mg/dL [IQR, 9-25] at baseline to 46 mg/dL [IQR, 21-71] on day 4 of therapy. Serum BUN levels increased significantly from baseline to discharge or urea discontinuation (18.4 ± 13.1 vs 41.1 ± 26.6; P ≤ .001). Urea was discontinued in 38 patients; the primary reason was due to palatability reported in 20 patients. Other reasons for discontinuation were serum sodium overcorrection (n = 5), serum sodium normalized (n = 5), ineffective response (n = 3), elevated BUN (n = 2), anticipated cost barrier upon discharge (n = 1), patient deceased (n = 1), and strict nothing by mouth (NPO) without medications (n = 1). All reasons for discontinuation were extracted, even if more than one was noted per patient.
Table 3.
Patients Experiencing Serum Sodium Overcorrection.
| Patient | Baseline-24 h | 24-48 h | 48-72 h | 72-96 h | ||||
|---|---|---|---|---|---|---|---|---|
| Na change (mEq/L) | Urea dose (g/day) | Na change (mEq/L) | Urea dose (g/day) | Na change (mEq/L) | Urea dose (g/day) | Na change (mEq/L) | Urea dose (g/day) | |
| 1 | −4 | 30 | 13 | 45 | — | — | — | — |
| 2 | 11 | 15 | 15 | 15 | — | — | — | — |
| 3 a | 6 | 30 | 9 | 60 | — | — | — | — |
| 4 b | 17 | 60 | 0 | 90 | 0 | 30 | 0 | 90 |
| 5 | 14 | 30 | — | — | — | — | — | — |
| 6 c | 12 | 15 | — | — | — | — | — | — |
Note. The entire study population (n = 74) was assessed for serum sodium overcorrection. Patients listed above include those who were on urea alone and those who were on urea with concomitant therapy. Patients who were on urea with concomitant therapy are noted in the footnotes below.
300 mEq Na from 0.9% sodium chloride infusion and sodium chloride tablets discontinued on day 1 of urea therapy.
Patient received higher than recommended starting dose.
4000 mEq Na received from 3% sodium chloride infusion and 0.9% sodium chloride infusion 1 day prior to urea.
A sub-analysis comparing patients receiving urea alone (n = 39) to those receiving urea with concomitant therapy (n = 35) was performed. No differences were observed between groups in mean baseline serum sodium, 24 hours, and at discharge or discontinuation (Table 4).
Table 4.
Serum Sodium (mEq/L) Urea Alone Versus Urea with Concomitant Therapy.
| Primary outcome | Urea alone | Urea with concomitant therapy | P |
|---|---|---|---|
| Baseline | 124 ± 4.1 | 124 ± 3.9 | .734 |
| 24 h | 126.6 ± 5 | 126.4 ± 4.4 | .855 |
| Discontinuation/discharge | 129 ± 4.3 | 131 ± 5.8 | .230 |
Note. Data are presented as mean ± standard deviation (SD).
Discussion
In this single-center retrospective cohort study, urea increased median serum sodium 2 mEq/L per day at a median dose of 30 g/day which is comparable to previous inpatient studies.4,9,11,12 Although our efficacy outcomes are similar to other published studies, safety outcomes including serum sodium overcorrection and patient intolerance have not been adequately described in the existing literature. Our findings establish the need for further evaluation of urea’s safety profile. Rondon-Berrios et al. 11 evaluated the role of urea in hyponatremia of any cause in 58 patients. Similar to our findings, patients received between 7.5 and 90 g of urea over a median of 4.5 days. Median serum sodium levels increased from a baseline of 124 mEq/L [IQR, 122-126] to 131 mEq/L [IQR, 127-134] at the end of therapy. In contrast to our findings, investigators in this previous study reported no incidence of serum sodium overcorrection or patient intolerance. We included a robust patient population compared to existing literature, including patients with mild, moderate, severe, hypovolemic, euvolemic, and hypervolemic hyponatremia. Though not directly assessed in this study, it is likely that our patient population also included patients with symptomatic, asymptomatic, acute, and chronic hyponatremia. It appears there was high prescriber acceptance of this new formulary agent at our institution as 39 (53%) patients received urea alone with no concomitant treatment for hyponatremia.
Serum sodium overcorrection occurred in 6 patients (8%) in our study and was the second most common reason for urea discontinuation. One patient experienced a significant serum sodium increase of 17 mEq/L in the first 24 hours following a urea dose of 60 g/day, which is the highest recommended starting dose per the package insert and higher than our institutional starting dose of 30 g/day. Although we did not correlate daily urea dose with magnitude of serum sodium increase, it is possible that dosing variability may contribute to serum sodium overcorrection in some patients. Many previous studies reported no incidence of serum sodium overcorrection which is possibly due to a less-strict definition of overcorrection when compared to our study; however, 5 of 6 patients in our sample experienced a serum sodium increase greater than 10 mEq/L in 24 hours.
Urea was used in 4 patients (5%) for hypovolemic hyponatremia in our study, which is a contraindication to its use.1,8 It was also discontinued in 5 patients (7%) once normal serum sodium was achieved. Depending on the underlying cause and chronicity of hyponatremia, this may not be a valid clinical practice. 1 SIADH was the most common cause of hyponatremia in our study and is an approved indication for urea. As a chronic disease state, therapy for SIADH should be continued to maintain normal serum sodium levels. It may be reasonable to discontinue urea upon serum sodium correction if the underlying cause for hyponatremia is determined to be an acute process; however, this practice has not been recognized in expert panel recommendations.1,8
Other factors that limit the use of urea are patient intolerance and outpatient affordability. Only 1 previous trial has reported on intolerance with discontinuation of therapy in 5 out of 36 patients (14%). 12 In our study, patient intolerance due to taste was the primary reason for discontinuation, affecting 20 patients (27%). Aside from clinical factors, cost may be a barrier when considering urea versus vasopressin antagonists in the outpatient setting due to the lack of insurance coverage and out-of-pocket expenses.
Our study had several limitations. First, the lack of a control group makes it difficult to claim full efficacy. It is also recommended to raise serum sodium concentration by 4 to 6 mEq/L in a 24-hour period1,8 while serum sodium in our study increased a median of 2 mEq/L per day. Though our results are statistically significant, it is difficult to determine the clinical significance given that serum sodium was increased at a slower rate than treatment guideline recommendations. Second, fluid restriction is a common treatment strategy for hyponatremia; however, we did not extract this data due to the variability and unreliability of intake and output charting. Third, although a high number of patients (42%) were discharged on urea per the discharge summary, we were unable to retrospectively assess insurance approval and prescription pick-up. It is possible that the number of patients who continued urea upon discharge is lower. Last, we observed several missing data points including serum sodium, and BUN; however, data were only collected while patients had an active order for urea, which was discontinued in greater than 50% of patients during hospitalization. Missing data points were excluded from the analysis.
To our knowledge, our study contains the largest sample size in the existing literature evaluating urea for treatment of hyponatremia of any cause. Our findings also add significant insight into the safety and palatability of urea which has not been adequately described to date. Given the adverse effect profile and monitoring considerations of urea, our results underscore the importance of clinical guidance and pharmacy oversight of this medical food which is a cost-effective inpatient alternative to vasopressin antagonists.
Conclusion
Although recommended as a second-line alternative agent in the 2013 U.S. expert panel and the 2014 European clinical guidelines,1,8 robust data supporting the use of urea is still lacking. This study demonstrated that urea is an effective treatment for hyponatremia in the inpatient setting; however, patient intolerance, serum sodium overcorrection, and affordability may limit its use in the outpatient setting. Further analysis on the safety profile of urea is needed to establish its use.
Footnotes
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iD: William M. Hammonds 
https://orcid.org/0000-0002-3654-4566
References
- 1. Verbalis JG, Goldsmith SR, Greenberg A, et al. Diagnosis, evaluation, and treatment of hyponatremia: expert panel recommendations. Am J Med. 2013;126(10):S1-S42. doi: 10.1016/j.amjmed.2013.07.006 [DOI] [PubMed] [Google Scholar]
- 2. Ure-Na. Prescribing information. Accessed November 2019. https://www.ure-na.com/articles.asp?ID=83
- 3. Otvos B, Kshettry VR, Benzel EC. The history of urea as a hyperosmolar agent to decrease brain swelling. Neurosurg Focus. 2014;36(4):E3. doi: 10.3171/2014.1.FOCUS13558 [DOI] [PubMed] [Google Scholar]
- 4. Decaux G, Genette F. Urea for long-term treatment of syndrome of inappropriate secretion of antidiuretic hormone. Br Med J (Clin Res Ed). 1981;283(6299):1081-1083. doi: 10.1136/bmj.283.6299.1081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Decaux G, Andres C, Gankam Kengne F, Soupart A. Treatment of euvolemic hyponatremia in the intensive care unit by urea. Crit Care. 2010;14(5):R184. Epub 2010 Oct 14. doi: 10.1186/cc9292. PMID: 20946646; PMCID: PMC3219290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Urea (Systemic). In: Lexicomp Online [proprietary Data]. UpToDate, Inc.; 2020. Accessed February 13, 2020. http://online.lexi.com/lco/action/home [Google Scholar]
- 7. Gankam Kengne F, Couturier BS, Soupart A, Decaux G. Urea minimizes brain complications following rapid correction of chronic hyponatremia compared with vasopressin antagonist or hypertonic saline. Kidney Int. 2015;87(2):323-331. doi: 10.1038/ki.2014.273 [DOI] [PubMed] [Google Scholar]
- 8. Spasovski G, Vanholder R, Allolio B, et al. Clinical practice guideline on diagnosis and treatment of hyponatraemia. Eur J Endocrinol. 2014;170(3):G1-G47. doi: 10.1530/eje-13-1020. Published 2014 Feb 25. [DOI] [PubMed] [Google Scholar]
- 9. Coussement J, Danguy C, Zouaoui-Boudjeltia K, et al. Treatment of the syndrome of inappropriate secretion of antidiuretic hormone with urea in critically ill patients. Am J Nephrol. 2012;35(3):265-270. doi: 10.1159/000336716 [DOI] [PubMed] [Google Scholar]
- 10. Soupart A, Coffernils M, Couturier B, Gankam-Kengne F, Decaux G. Efficacy and tolerance of urea compared with vaptans for long-term treatment of patients with SIADH. Clin J Am Soc Nephrol. 2012;7(5):742-747. doi: 10.2215/CJN.06990711 [DOI] [PubMed] [Google Scholar]
- 11. Rondon-Berrios H, Tandukar S, Mor MK, et al. Urea for the treatment of hyponatremia. Clin J Am Soc Nephrol. 2018; 13(11):1627-1632. doi: 10.2215/CJN.04020318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Nervo A, D’Angelo V, Rosso D, et al. Urea in cancer patients with chronic SIAD-induced hyponatremia: old drug, new evidence. Clin Endocrinol (Oxf). 2019;90(6):842-848. doi: 10.1111/cen.13966 [DOI] [PubMed] [Google Scholar]
- 13. Tolvaptan. In: Lexicomp Online [proprietary Data]. UpToDate, Inc. 2020. Accessed March 3, 2020. http://online.lexi.com/lco/action/home [Google Scholar]
- 14. Conivaptan. In: Lexicomp Online [proprietary Data]. UpToDate, Inc. 2020. Accessed May 8, 2020. http://online.lexi.com/lco/action/home [Google Scholar]
- 15. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377-381. doi: 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Harris PA, Taylor R, Minor BL, et al. The REDCap consortium: building an international community of software platform partners. J Biomed Inform. 2019;95(April):103208. doi: 10.1016/j.jbi.2019.103208 [DOI] [PMC free article] [PubMed] [Google Scholar]