LABORATORIO DE URGENCIAS

Abbreviated Urine Collection Compared With 24-Hour Urine Collection for Measuring Creatinine Clearance in Adult Critically Ill Patients.

          Mohamed Omar Saad, PharmD, BCPS; Adham Mohamed, PharmD, BCPS and Mohamed Izham Mohamed Ibrahim, PhD .

                            Annals of Pharmacotherapy 1–10 (2024) - DOI: 10.1177/10600280241241820 (Systematic Review)

Dose adjustment of renally-excreted drugs is necessary to prevent drug accumulation and toxicities. Appropriate renal drug dosing is particularly important in critically ill patients because of their higher susceptibility to adverse drug effects.1 In noncritically ill adult patients, drug-dosing can be guided by formulas that estimate creatinine clearance (CrCl) or glomerular filtration rate (GFR) such as Cockcroft[1]Gault (CG) equation, the Modification of Diet in Renal Disease (MDRD) equation, and the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation.2-5 The estimated CrCl or GFR is then used to select the appropriate dosing regimens for specific drugs.2
The usefulness of the previously mentioned formulas in critically ill patients is uncertain. These formulas were orig[1]inally developed in noncritically ill patients.2-5 In addition, they assume steady-state serum creatinine, but most criti[1]cally ill patients have unstable serum creatinine and volume status that alter the pharmacokinetics of creatinine.2 Several studies have demonstrated that estimating formulas have high bias and poor precision in critically ill patients, com[1]pared with CrCl measured using 24-hour urine collection6 and compared with the use of exogenous markers such as chromium-ethylenediaminetetraacetic acid (51Cr-EDTA)7 and inulin8 . Measuring CrCl using 24-hour urine collection provides more accurate quantification of renal function than estima[1]tion formulas.2 However, the long duration of urine collec[1]tion makes it error-prone and labor-intensive.9 It is also impractical to delay dose adjustments in critical illness until 24-hour urine collection is completed, in addition to the time required by the laboratory to report creatinine values. Shorter durations of urine collection may provide a faster alternative to guide timely renal dose adjustments in criti[1]cally ill patients. In this systematic review, we aimed to evaluate the accu[1]racy of abbreviated urine collection compared with 24-hour urine collection for measuring CrCl in critically ill adult patients

METHODS

* Search Strategy and Selection Criteria
This systematic review followed the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) Statement,10 and targeted the following PICO question: in critically ill adult patients, what is the accuracy of creati[1]nine clearance measured by abbreviated urine collection (≤12 hours) compared with 24-hour urine collection. Eligible studies were comparative studies, including ran[1]domized controlled trials, nonrandomized controlled trials, before and after studies, prospective and retrospective cohort studies, case-control studies, and analytical cross[1]sectional studies. Studies were included if they involved adult critically ill patients admitted to any type of ICU, for example, medical or surgical. We excluded descriptive observational study designs and qualitative studies. Reviews and opinion papers were not included, but their reference lists were screened for eligible studies. We undertook an initial search of PubMed to identify relevant articles on the topic. We used the text words con[1]tained in the titles and abstracts of relevant articles and the index terms describing relevant articles to develop a complete search strategy. We searched PubMed, Embase, Web of Science, and Google Scholar using the complete search strategy. We also searched ProQuest Dissertations and Thesis Global (available at https://www.proquest. com). Full details of the search strategy are available in Appendix A. We ran the searches on March 4, 2023. We searched the databases from the date of inception (with[1]out any limitation on the dates) and included studies pub[1]lished in English only. We also screened the reference lists of included studies for eligible studies.
We updated the search on March 1, 2024, using the same search strat[1]egy, and filtered the results to include only publication dates on or after March 4, 2023 (the date of initial search), and did not find more relevant records by title/abstract screening.
* Selection of Studies and Data Extraction
Following the search, all identified citations were collated and uploaded into EndNote 20 (Clarivate Analytics, PA, USA), and duplicates were removed. After duplicate removal, records were uploaded to the Rayyan online tool; and titles and abstracts were screened independently by 2 reviewers (MS and AM). Disagreements were resolved by discussion and consensus.
Potentially relevant records were sought for retrieval in full. The full text of identified records was assessed in detail against the inclusion criteria by 2 reviewers (MS and AM) independently.
Reasons for exclusion of full[1]text studies that do not meet the inclusion criteria were recorded. The corresponding authors were contacted to pro[1]vide the full text of their studies when they were not available online.
The search results and the study inclusion process were reported in full and presented in a PRISMA flow diagram.10
* Data Extraction
Two reviewers (MS and AM) independently extracted rele[1]vant data from the included studies using a customized Microsoft Excel data extraction form developed by the reviewers. Discrepancies in the extracted data were resolved by consensus after revisiting the full-text articles.  The data extracted included aim(s), design, details of participants, settings, renal function quantification methods, analysis methods, and outcome measures. We contacted the corre[1]sponding authors of included studies to request missing or additional data when required
* Outcome Measures
To evaluate the agreement between CrCl measured by abbreviated collection and 24-hour urine collection, mea[1]surement bias, estimated as the mean of differences between the 2 methods (mean prediction error), was extracted from the included studies. The root of mean squared prediction error was sought as a measure of precision as described by Sheiner and Beal.11 Saad et al 3 Correlation coefficients were extracted from the included studies as well. Where coefficients of determination were reported from simple linear regression of measured 24-hour CrCl on abbreviated CrCl, correlation coefficients (r) were calculated as the square root of coefficients of determina[1]tion (R2 )

 RESULTS

* Study Selection
A total of 5758 unique records were identified through database searching. Title and abstract screening yielded 11 records, of which 4 abstracts did not have full text (3 stud[1]ies published as conference abstracts13-15, which did not report comparing abbreviated CrCl to 24-hour CrCl; and 1 study published in 1987,16 which did not have full text available online). We contacted the corresponding authors of these 4 abstracts twice by e-mail to get the full text, but we did not get a reply. The remaining 7 records underwent full-text review. Of those, 1 article was in Chinese,17 1 did not assess the accuracy of absolute CrCl measures but only assessed the change in CrCl as a marker of acute kid[1]ney injury,18 and 1 was of noncritically ill patients.19 Five records were identified from sources other than database searching; 3 of them were excluded: 2 studies compared CrCl using abbreviated urine collection to clear[1]ance of exogenous substances (51Cr-EDTA and cystatin C) and the reported data were insufficient to establish direct comparison to 24-hour CrCl,7,8 and 1 used 1-hour CrCl as a reference group to which CrCl estimation formulas were compared.20 In addition, 1 record was for a conference abstract of a study identified through the database search.21
Overall, we included 4 studies from database search[1]ing,9,22-24 and 1 study identified through other sources.25
* Description of Included Studies
The 5 studies comprising 528 adult critically ill patients were published between 1980 and 2007. Four studies were conducted in the United States22-25 and 1 in Spain.9 The study design was stated to be prospective in 4 studies9,22,24 and was not explicitly stated in 1 study,25 which we classi[1]fied as prospective based on the described study methods. The studies were conducted in different types of ICUs (medical, surgical, trauma/surgical, and mixed). The num[1]ber of critically ill patients included in each study ranged from 19 to 359. Three studies excluded patients with oligu[1]ria,23-25 1 excluded any patient with anuria,9 and 1 excluded patients with end-stage renal disease who were anuric.22 Two studies reported using more than 1 sample set for indi[1]vidual patients.22,25 The mean 24-hour measured CrCl ranged from 57 mL/ min/1.73 m2 to 103 mL/min. Three studies used 24-hour CrCl as a reference group,9,22,24 1 used 22-hour CrCl,25 and 1 used inulin clearance.23
Three studies evaluated the CG formula for estimating CrCl in addition to abbreviated urine collection.9,22,23 Table 1 shows the detailed characteristics of the included studies. Risk of Bias Three studies were judged using the QUADAS-2 tool to be at high risk of bias22,24,25 and 2 studies were at low risk.9,23 Two studies were judged to have high concerns about appli[1]cability.22,24 Figure 2 summarizes the assessment of bias and applicability of included studies. The details of the risk of bias per domain and their justifications are presented in Appendix B
* Accuracy of Abbreviated Urine
Collection Different accuracy measures were used in the included studies (Table 1). Two studies reported bias (mean predic[1]tion error),9,24 and we calculated it from the reported data in 1 study.23 The bias ranged from 10 to 16 mL/min for differ[1]ent urine collection durations (reported as 16% to 25% of the 24-hour CrCl) in 1 study24, was 5 mL/min/1.73 m2 (9% of the mean 24-hour CrCl) in another study23 and was 8.3 mL/min/1.73 m2 (8% of the mean 24-hour CrCl) in the third study.9 Bias increased at higher ranges of mean 24-hour CrCl.9,24 Table 2 depicts the values of bias comparing abbre[1]viated CrCl to the reference group in each study. We calculated the precision from the data reported in 1 study.23 The mean of squared prediction error was 927.5 mL2 /min2 /1.73m4 , equivalent to a root mean squared prediction error of 30.5 mL/min/1.73 m2 . Only 1 study used the Bland-Altman method26 to assess agreement between measured CrCl using abbreviated urine collection and 24-hour CrCl.9 Two studies used linear regression,22,23 2 used correla[1]tion coefficients,9,24 and 1 used both25 to describe the rela[1]tionship between measured CrCl using abbreviated urine collection and 24-hour urine collection. Correlation coeffi[1]cients ranged from 0.8 to 0.95 and were generally higher for longer durations of urine collection (Table 2). One study reported that the correlation between the 2-hour and 24-hour CrCl was particularly good in patients with 2-hour CrCl of less than 80 mL/min and was lowest in 6 patients with CrCl above 120 mL/min.24

DISCUSSION

In this systematic review, we found that CrCl measured using abbreviated urine collection (≤ 12 hours) correlated well with CrCl measured using 24-hour urine collection. However, it has a measurement bias ranging from 5 mL/ min/1.73 m2 to 16 mL/min (from 8% to 25% of the 24-hour CrCl). In addition, abbreviated urine collection led to an imprecise measurement of CrCl in 1 study (root mean squared prediction error = 30.5 mL/min/1.73 m2 ). Thus, using abbreviated urine collection may overestimate CrCl and result in dosing recommendations higher than those based on 24-hour CrCl. Measurement of CrCl using 24-hour urine collection requires accurate urine collection for a relatively long time, which increases the chances of error, increases the workload on bedside nurses, and makes it unsuitable for repeated mea[1]surements.9 In addition, the time required for urine collec[1]tion and sample analysis in the laboratory limits its usefulness for timely dose adjustments, which is crucial in critically ill patients. The use of abbreviated urine collection has emerged as a potential alternative to overcome these limitations. Indeed, one of the included studies reported that 24-hour measurements could not be obtained in 13.9% of the study population, whereas only 0.6% of the 2-hour measurements could not be obtained.9
Furthermore, several studies have used CrCl calculated using abbreviated urine collection as a reference group to which other methods of CrCl estimation were compared.14,15,20 However, given the findings of this review, abbreviated CrCl should be used with caution until more robust evidence is available.
Regarding the assessment of included studies, using 24-hour CrCl as a reference method, instead of directly mea[1]sured GFR, was not considered a source of bias. Although directly measured GFR using exogenous substances such as inulin and 51Cr-EDTA as filtration markers represents the gold standard for assessing renal function,9 its use is limited to research purposes but is not applicable to day-to-day practice.2 2 In addition, abbreviated urine collection is not meant to reproduce the GFR values obtained from exogenous sub[1]stances but to substitute the 24-hour CrCl. Thus, 24-hour CrCl was considered an acceptable reference group in the included studies as it represents the practical reference method. Isotope dilution mass spectrometry (IDSM) methods for analysis of creatinine levels provide more accurate mea[1]surements than the old non-IDMS methods.27 Some included studies were conducted before the widespread use of IDMS methods. However, non-IDMS analysis is unlikely to have affected the results of these studies because devia[1]tions from the true values would have impacted both serum and urine levels and would have canceled out in the calcula[1]tion of measured creatinine clearance.24,25 Thus, using non[1]IDMS methods for measuring creatinine was not considered a high risk of bias in the included studies. Glomerular filtration rate varies throughout the day.28 Two included studies reported measuring abbreviated CrCl at specific times of the day for all patients to minimize diur[1]nal variability.22,25 In addition, Wilson et al reported that the time of urine collection has no significant impact on the agreement between 2-hour CrCl and 24-hour CrCl in 30 critically ill patients.24 However, the impact of diurnal vari[1]ability of GFR and urine collection time on the accuracy of abbreviated CrCl needs to be re-evaluated in larger studies.
The studies included in the current review have several limitations. None of the included studies differentiated between patients with stable renal function and patients with acute kidney injury. Two studies included multiple samples for individual patients,22,25 which may have introduced bias to the results because of the correlation between samples collected from the same patient. In addition, 3 studies23-25 excluded patients with oliguria, limiting the generalizability of their results to patients with this common finding in the ICU. Finally, some of the included studies have limited applicability in contemporary practice as they were pub[1]lished a long time ago and represent different clinical set[1]tings. Fluid management in the ICU, for example, has changed over the years,29 which may substantially impact the values of measured CrCl. Further studies that avoid the previous limitations are needed to establish the role of abbre[1]viated CrCl for drug dosing in adult critically ill patients. To our knowledge, this is the first review to assess the use of abbreviated urine collection for measuring CrCl in adult critically ill patients. However, it has some limita[1]tions. We could not retrieve 4 abstracts (3 studies published as conference abstracts and 1 old study, which did not have full text available online). We contacted the corresponding authors to get the full text, but we did not receive a reply. However, the 3 conference abstracts did not compare abbre[1]viated urine collection to 24-hour CrCl, and it is unlikely that the full text would have such a comparison. In addition, we excluded 1 study published in Chinese due to the inabil[1]ity to extract the accurate details of the study even after translation to English. This excluded study might have included findings relevant to our review.
* Relevance to Patient Care and Clinical
Abbreviated urine collection (≤12 hours) is commonly used as an alternative to 24-hour urine collection to measure CrCl for renal drug dosing in critically ill patients. Inaccurate mea[1]surement of CrCl may lead to inappropriately high or low doses of renally excreted medications, which exposes criti[1]cally ill patients to a higher risk of adverse events or therapeu[1]tic failure, respectively. This review evaluated the accuracy of abbreviated urine collection compared with the 24-hour col[1]lection for measuring CrCl in critically ill adult patients.

CONCLUSION

The evidence on the usefulness of abbreviated urine collec[1]tion for measuring CrCl is scarce and of limited quality. This evidence suggests that measured CrCl using abbrevi[1]ated urine collection does not adequately replace 24-hour CrCl due to measurement bias and imprecision. Using abbreviated urine collection for CrCl measurement may overestimate CrCl, resulting in dosing recommendations above those based on 24-hour CrCl. Large prospective high-quality studies are needed to re-evaluate the role of abbreviated urine collection in measuring CrCl. Such stud[1]ies should utilize measures of bias and precision to compare CrCl measured using different durations of abbreviated urine collection to 24-hour CrCl.

NOTE: tables, graphs, charts in the bibliography cited at the beginning

REFERENCES

1. Smith BS, Yogaratnam D, Levasseur-Franklin KE, Forni A, Fong J. Introduction to drug pharmacokinetics in the criti[1]cally ill patient. Chest. 2012;141(5):1327-1336. doi:10.1378/ chest.11-1396
   2. Matzke GR, Aronoff GR, Atkinson AJ Jr, et al. Drug dos[1]ing consideration in patients with acute and chronic kidney disease—a clinical update from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney Int. 2011;80(11):1122- 1137. doi:10.1038/ki.2011.322
   3. Levey AS, Stevens LA, Schmid CH, et al. A new equa[1]tion to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. doi:10.7326/0003-4819-150-9-200905 050-00006 4. Levey AS, Coresh J, Greene T, et al. Expressing the Modification of Diet in Renal Disease Study equation for esti[1]mating glomerular filtration rate with standardized serum cre[1]atinine values. Clin Chem. 2007;53(4):766-772. doi:10.1373/ clinchem.2006.077180
   5. Cockcroft DW, Gault MH. Prediction of creatinine clear[1]ance from serum creatinine. Nephron. 1976;16(1):31-41. doi:10.1159/000180580
   6. Al-Dorzi HM, Alsadhan AA, Almozaini AS, et al. The per[1]formance of equations that estimate glomerular filtration rate against measured urinary creatinine clearance in criti[1]cally ill patients. Crit Care Res Pract. 2021;2021:5520653. doi:10.1155/2021/5520653
   7. Bragadottir G, Redfors B, Ricksten SE. Assessing glomerular filtration rate (GFR) in critically ill patients with acute kid[1]ney injury—true GFR versus urinary creatinine clearance and estimating equations. Crit Care Lond Engl. 2013;17(3):R108. doi:10.1186/cc12777
   8. Carlier M, Dumoulin A, Janssen A, et al. Comparison of different equations to assess glomerular filtration in criti[1]cally ill patients. Intensive Care Med. 2015;41(3):427-435. doi:10.1007/s00134-014-3641-9
   9. Herrera-Gutiérrez ME, Seller-Pérez G, Banderas-Bravo E, Muñoz-Bono J, Lebrón-Gallardo M, Fernandez-Ortega JF. Replacement of 24-h creatinine clearance by 2-h creatinine clearance in intensive care unit patients: a single-center study. Intensive Care Med. 2007;33(11):1900-1906. doi:10.1007/ s00134-007-0745-5
   10. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372: n71. doi:10.1136/bmj.n71
   11. Sheiner LB, Beal SL. Some suggestions for measuring predic[1]tive performance. J Pharmacokinet Biopharm. 1981;9(4):503- 512. doi:10.1007/BF01060893
   12. Whiting PF. QUADAS-2: a revised tool for the qual[1]ity assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536. doi:10.7326/0003-4819-155-8- 201110180-00009
   13. Janssen A, Dumoulin A, De Waele JJ, Vanholder R, Waterloos MA, Hoste EA. Assessment of increased glo[1]merular filtration rate in ICU patients. Intensive Care Med. 2010;36(23):S402. 14. Fleuren L, Lemmens H, de Grooth HJ, et al. Right dose, right now: volume-based urine collection to assess creatinine clear[1]ance in the critically ill. Intensive Care Med Exp. 2018;6:40. doi:10.1186/s40635-018-0201-6
   15. Rajeevkumar P, Ostermann M, Hanks F. Utility of 8 h urine creatinine clearance to guide dosing in critically ill patients: a single centre retrospective analysis. Crit Care. 2022;26:P133. doi:10.1186/s13054-022-03927-z
   16. Baumann TJ, Staddon JE, Horst HM, Bivins BA. Minimum urine collection periods for accurate determination of cre[1]atinine clearance in critically ill patients. Clin Pharm. 1987;6(5):393-398.
   17. Jia LJ, Jiang L. [A comparative study of 2- and 24-hour cre[1]atinine clearance in patients in intensive care unit]. Zhonghua Yi Xue Za Zhi. 2018;98(47):3878-3882. doi:10.3760/cma.j.i ssn.0376-2491.2018.47.013
   18. Pickering JW, Frampton CM, Walker RJ, Shaw GM, Endre ZH. Four hour creatinine clearance is better than plasma cre[1]atinine for monitoring renal function in critically ill patients. Crit Care. 2012;16(3):R107. doi:10.1186/cc11391
   19. Herget-Rosenthal S, Kribben A, Pietruck F, Ross B, Philipp T. Two by two hour creatinine clearance—repeatable and valid. Clin Nephrol. 1999;51(6):348-354.
   20. Hoste EA, Damen J, Vanholder RC, et al. Assessment of renal function in recently admitted critically ill patients with normal serum creatinine. Nephrol Dial Transplant. 2005;20(4):747- 753. doi:10.1093/ndt/gfh707
   21. Seller-Pérez G, Herrera-Gutiérrez ME, Banderas-Bravo E, Muñoz-Bono J, Fernández-Ortega J, Lebrón-Gallardo M. Validation of 2 hours sampling creatinine clearance in ICU population, included unstable patients. Intensive Care Med. 2005;31(1)(suppl):S146.
   22. Cherry RA, Eachempati SR, Hydo L, Barie PS. Accuracy of short-duration creatinine clearance determinations in predict[1]ing 24-hour creatinine clearance in critically ill and injured patients. J Trauma. 2002;53(2):267-271. doi:10.1097/00005 373-200208000-00013
   23. Robert S, Zarowitz BJ, Peterson EL, Dumler F. Predictability of creatinine clearance estimates in critically ill patients. Crit Care Med. 1993;21(10):1487-1495. doi:10.1097/00003246- 199310000-00016
   24. Wilson RF, Soullier G. The validity of two-hour cre[1]atinine clearance studies in critically ill patients. Crit Care Med. 1980;8(5):281-284. doi:10.1097/00003246- 198005000-00002 10 Annals of Pharmacotherapy 00(0
   25. Sladen RN, Endo E, Harrison T. Two-hour versus 22-hour creatinine clearance in critically ill patients. Anesthesiology. 1987;67(6):1013-1016. doi:10.1097/00000542-198712000- 00031 26. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet Lond Engl. 1986;1(8476):307-310.
   27. Myers GL, Miller WG, Coresh J, et al. Recommendations for improving serum creatinine measurement: a report from the laboratory working group of the National Kidney Disease Education Program. Clin Chem. 2006;52(1):5-18. Accessed May 17, 2023. https://academic.oup.com/clinchem/arti[1]cle/52/1/5/5626617
   28. Wesson LG Jr, Lauler DP. Diurnal cycle of glomerular filtra[1]tion rate and sodium and chloride excretion during responses to altered salt and water balance in man. J Clin Invest. 1961;40(11):1967-1977.
   29. Finfer S, Myburgh J, Bellomo R. Intravenous fluid therapy in critically ill adults. Nat Rev Nephrol. 2018;14(9):541-557. doi:10.1038/s41581-018-0044-0