Skip to main content

Advertisement

Table 1 Trial Summary Table by Study Type (n=36)

From: The impact of “early” versus “late” initiation of renal replacement therapy in critical care patients with acute kidney injury: a systematic review and evidence synthesis

Author, Year Study Design Country Duration Exclusion Patient Population Patients (n) Age (mean) yrs Illness Severity Score Early RRT Criteria Late RRT Criteria Study Quality Primary Outcome
Total Early RRT Late RRT
Randomized Trials
Bouman, 2002 [12] RCT, two-center study Netherlands May 1998 - Mar 2000 Pre-existing renal disease Multisystem 106 70 36 EHV: 68; ELV: 70; LLV: 67 EHV: SOFA 10.3 - APACHE2=23.5, ELV: SOFA 10.1 - APACHE2=21.7; LLV: SOFA 10.6 - APACHE2=23.6 TIME: Early < 12 h (200ml); Early Low Vol < 12 h (100-150ml) TIME: Late > 12h HIGH 28 d mortality: EHV: 9/35(26%) died, ELV: 11/35(31%) died, LLV: 9/36(25%) died; p=0.8
Durmaz, 2003 [13] RCT Turkey Sept 1999 - Aug 2001 Age<18, chronic dialysis Post Cardiac Surgery 44 21 23 Early 58; Late 54 NR BIOCHEM: Cr rise >10% from pre-op level within 48hrs of surgery Cr rise >50% from pre-op level; or Urine output <400ml/24hrs with coexistent K+/H+ unresponsive to med mgmt LOW Hospital mortality: Early 1/21 (4.8%) died, Late 7/23 (30.4%) died p=0.048; Favors Early
Sugahara, 2004 [14] RCT Japan Jan 1995 - Dec 1997 Pregnancy, Bili > 5mg/dL, Mental disorder, Cancer, Early recovery of urine output >30ml/kg/hr prior to RRT Post Cardiac Surgery 28 14 14 Early: 65; Late: 64 Early: APACHE2=19; Late: APACHE2=18 BIOCHEM: UOP <30ml/hr × 3hrs OR UOP <750ml/day; Mean time to RRT start 18d±0.9 post op UOP<20ml/hr × 2hrs+ OR UOP <500ml/day; Mean time to RRT start 1.7d±0.8 post op HIGH 14 d mortality: Early 2/14 died (14%), Late 12/14 died (86%); p<0.01 Favors Early
Payen, 2009 [7] RCT, multicenter France Jan 1997 - Jan 2000 Age<18, chronic dialysis, pregnant, moribund state, prior immunosuppressive therapy Multisystem 76 37 39 Early 58 Late 59 Early: SOFA 11.6- SAPS2 54.3; Late: SOFA 10.4- SAPS2 52.4 TIME: Protocolized RRT × 96hrs w/ diagnosis of ‘sepsis’. Mean time to initiation of RRT not specified Control = No RRT unless metabolic renal failure & classic indications for RRT present HIGH Early 20/37 (54%) died, Late 17/37 (44%) died; p = 0.49
Jamale, 2013 [15] RCT, single center India April 2010 - July 2012 Required urgent dialysis at time of randomization Multisystem 208 102 106 Early 43 Late 42 Early: SOFA 7.3; Late: SOFA 8.2 BIOCHEM: Cr > 618μmol/L Classic indications for RRT, Symptomatic uremia unresponsive to med mgmt HIGH Mortality: Early 21/102 (20.5%) died, Late 13/106 (12%); p=0.2
Combes, 2015 [16] RCT, multicenter USA 2009-2012 <18, Pregnant, Chronic RRT, Weight >120kg, SAPS II>90 (i.e. moribund) Post Cardiac Surgery 224 112 112 Early 61 Late 58 Early: SOFA 11.5- SAPS2=54; Late: SOFA 12.0- SAPS2=55.1 TIME: RRT initiated <24hrs and continued for min of 48hrs; Mean time to randomization 12hrs Classic indications for RRT, Lifethreatening metabolic derangements unresponsive to med mgmt HIGH Mortality: Early 40/112 (36%) died, Late 40/112 (36%) died; p = 1.0
Wald, 2015 [17] RCT, multicenter Canada May 2012 - Nov 2013 Intoxication requiring RRT, Limited resuscitation directives, RRT within the previous 2 months, RPGN, Obstructive uropathy, > 48hrs to doubling time of Cr Multisystem 100 48 52 Early 62 Late 64 Early: SOFA 13.3 Late: SOFA 12.8 TIME: Time from randomization < 12h; Mean time to RRT = 9.7hrs Intensivist judgement regarding hyperkalemia, volume overload, acidemia refractory to medical therapy, Uremic symptoms Mean time to RRT=32hrs HIGH Mortality: Early 16/48 (33%) died, Late 19/52 died; p = 0.74
RCT Totals       786 404 382       Pooled mortality: Early 120/404 (29.7%), Late 117/382 (30.6%); n=7
Prospective Trials
Liu, 2006 [18] Prospective Observational Multicentre Multi countries Feb 1999 - Aug 2001 GFR<30ml/min/1.73m2 Multisystem 243 122 121 Early 54 Late 58 NR Azotemia defined by BUN<76mg/dL Azotemia defined by BUN>76mg/dL LOW NOQA=6 28 d mortality: Early 43/122(35%) died vs Late 50/121(41%) P=0.09 Favors Early
Iyem, 2009 [19] Prospective Observational cohort Turkey May 2004 - April 2007 Preexisting renal disease and pre operative high levels of urea and creatinine Post cardiac surgery 185 95 90 Early: 64; Late: 62 NR TIME: Evidence of 50% increase in BUN, low urine output (<0.5mL/kg/h) triggering RRT started < 48hrs TIME > 48hrs to start of RRT for similar markers of renal failure managed medically for minimum 48hrs LOW NOQA=7 In hosp mortality: Early 5/95(5%) died, Late 6/90(7%) died; NS
Bagshaw, 2009 [20] Prospective Observational Multicentre (BEST Kidney) 23 countries Sept 2000 - Dec 2001 Pre existing chronic RRT, drug toxicity, age <12 Multisystem 1227 959 268 Early: 60, Delayed: 63, Late: 64; p=0.003 Early: SOFA 10.9- SAPS2=53.5 Delayed: SOFA 11.1- SAPS2=46 Late: SOFA 10.7- SAPS2=43.1; p=0.04 TIME: Early RRT started for azotemia (Urea>30mmol/L or low urine output × 12h) <2d (n=785), Delayed RRT started 2-5d (n=174) from ICU admission RRT started >5d from ICU admission LOW NOQA=7 Hosp mortality: Early 462/785(59%) died, Delayed 108/174(62%) died, Late 195/268(72%) died; P<0.0011 Favors Early
Shiao, 2009 [21] Prospective Observational Multicentre Taiwan Jan 2002 - Dec2005 Prior dialysis, without surgery, or surgery did not involve abdominal cavity. History of renal trasplant Major abdominal surgery 98 51 47 Early: 65; Late: 68 Early: SOFA 8.3- APACHE2=18.2; Late: SOFA 8.5- APACHE2=18.8 BIOCHEM: RIFLE criteria: RISK or pre-RISK criteria (Mean Time to RRT from ICU Admit = 7.3d) RIFLE criteria: INJURY or FAILURE criteria (Mean Time to RRT from ICU Admit=8.4d) HIGH NOQA=7 Hosp mortality: Early 22/51(43%), Late35/47(75%); p=0.0028 Favors Early
Sabater, 2009 [22] Prospective Observational Spain 2 years NR Multisystem 148 44 104 All patients mean = 60; NR Early: APACHE2=26; Late: APACHE2=24 BIOCHEM: RRT initiated for RIFLE: RISK & INJURY; (Mean RRT start 2.2d post ICU admit) RRT initiated for RIFLE: FAILURE; (Mean RRT start 6.4d post ICU admit) LOW NOQA=7 Mortality: Early 21/44 died, Late 68/104 died. P=0.047 Favors Early
Elseviers, 2010 [23] Prospective Observational Multicentre Belgium 2001-2005 Pre existing renal disease (Cr<1.5mg/dl), reduced kidney size on ultrasound Multisystem 1303 653 650 Early 64; Late 67 Early: SOFA 9.9- APACHE2=25.2; Late: SOFA 8.5- APACHE2=5.2, p=0.001 BIOCHEM: Unspecified SHARF scoring criteria w/serum Cr > 2mg/dL Conservative approach = No RRT LOW NOQA=5 Mortality: Early 379/653 (58%) died, Late 280/650 (43%) died; p<0.001 Favors Late
Vaara, 2012 [24] Prospective Observational Multicentre (FINNAKI Study) Finland Sep 2011 - Feb 2012 NR Sepsis, Cardiogenic Shock 261 NR NR NR Survivors: SAPS2=47; Non-survivors: SAPS2=66 TIME: Time<24hrs from ICU admit Time> 24hrs from ICU admit LOW NOQA=5 OR for late 2.69 (1.07-6.73, p=0.035). Favors Early
Perez, 2012 [25] Prospective Observational Spain   NR Sepsis 244 135 109 Early 62; Late 62 Early: SOFA 12; Late: SOFA 11 TIME: Time from ICU admission to RRT < 48h TIME >48hrs LOW NOQA=5 90 d mortality: Early 71/135(53%) died, Late 78/109(72%) died; p=0.003. Favors Early
Lim, 2014 [27] Single Centre Prospective Cohort Singapore Dec 2010 - April 2013 Chronic dialysis patients, Dialysis initiated prior to ICU admission Medical & Surgical patients 140 84 56 Early 60; Late 64 Early: SOFA 7; Late: SOFA 11; p=0.001 BIOCHEM: AKIN stage 1 or 2 AND compelling indication or AKIN stage 3 (Cr≥354μmol/l or Cr>300% baseline w/urine <0.3cc/kg/h for 24h or anuria >12h) Traditional indications: K>6mmol/L, Urea ≥30mmol/L, pH<7.25, Bicarb <10mmol/L, Pulm edema, Uremic encephalopathy/pericarditis LOW NOQA=6 Hosp mortality: Early 36/84(43%) died, Late 37/56(66%) died; p=0.007 Favors Early
Jun, 2014 [26] Nested Observational, Multi-Centre Study ‘RENAL’ Study Group NZ, Australia Dec 2005 - Nov 2008 Age<18, Prior RRTduring admission, Prior RRT for CKD Sepsis 439 219 220 Early 65; Late 64 Early: SOFA: 2.0- APACHE3=107, Late: SOFA 2.1- APACHE3=100, P<0.001 TIME: AKI diagnosis to randomization < 17.6 hrs Time from AKI diagnosis to randomization >17.6hrs LOW NOQA=6 28 d mortality: Early 82/219(37%) died; Late 84/220(38%) died (p=0.923) NS
PROSPECTIVE TOTALS       4288 2362 1665       Pooled mortality: Early 1229/2362 (52%), Late 833/1665 (50%); n=10
Retrospective Trials
Gettings, 1999 [28] Retrospective cohort USA 1989 - 1997 CRRT duration <48hrs, Pediatric patients, Incomplete records Trauma 100 41 59 Early 40; Late 48 Early ISS = 33.0; Late ISS = 37.2 BIOCHEM: BUN < 60mg/dL AND Oliguria, Vol overload, Electrolytes, Uremia; Mean RRT start post admission day 10; p<0.0001 BUN > 60 mg/dL AND Oliguria, Vol overload, Electrolytes, Uremia; Mean RRT start post admission day 19 LOW NOQA=5 Hosp mortality: Early 25/41(61%) died, Late 47/59(80%) died; p=0.041 Favors Early
Elahi, 2004 [29] Retrospective cohort UK Jan 2002 - Jan 2003 Preexisting renal disease Post cardiac surgery 64 36 28 Early 69; Late 68 NR BIOCHEM: Low urine output = less than 100 ml within 8h after surgery;Mean RRT start 0.78 days Traditional indications: Urea ≥30mmol/L, Cr Elahi, 2004 [29] ≥250mmol/L, K > 6.0mEq/L; Mean RRT start 2.5 days LOW NOQA=6 28 d mortality: Early-8/36 died (22%), Late-12/28 (43%); p<0.05 Favors Early
Demirkilic, 2004 [30] Retrospective cohort Turkey Mar 1992 - Sep 2001 NR Post Cardiac Surgery 61 34 27 NR p=0.3 NR BIOCHEM: Low urine output = less than 100ml within 8hrs post op; Mean RRT start 0.88 days Cr≥5mg/dL, or K>5.5 mEq/L w/med mgmt; Mar 92-Jun 96; Mean RRT start 2.56 days LOW NOQA=6 Hosp mortality: Early 8/34(23%), Late 15/27(56%); P=0.016 Favors Early
Wu, 2007 [32] Retrospective cohort Taiwan July 2002- Jan2005 Hepatorenal syndrome from cirrhosis, liver trasplant, cardiopolmunary resuccitation Acute liver failure 80 54 26 Early 55; Late 63; p=0.03 Early: SOFA 12.4- APACHE2=18.2; Late: SOFA 13.2- APACHE2=20.5 BIOCHEM: BUN < 80 mg/dL AND traditional indications present Traditional indications present with BUN > 80mg/dL LOW NOQA=6 30 d mortality: Early 34/54(63%) died vs Late 22/26(85%) died; P=0.04 Favors Early
Andrade, 2007 [31] Retrospective cohort Brazil 2002-2005 Patients who did not have both AKI and respiratory failure believed secondary to leptospirosis Leptospirosis 33 18 15 Early 42; Late 44 Early: APACHE2=24.5; Late: APACHE2=26 TIME: Mean time to RRT = 265 min Mean time to RRT = 1638 min LOW NOQA=5 Hosp mortality: Early 3/18(17%) died, Late 10/15(67%) died; P=0.01 Favors Early
Manche, 2008 [33] Retrospective cohort Malta 1995-2006 NR Post Cardiac Surgery 71 56 15 Early 66; Late 63 NR BIOCHEM: Urine output<0.5ml/kg/hr unresponsive to med mgmt; Mean RRT start 8.6hrs post-op Oliguria (output < 0.5ml/Kg/hr) refractory to med mgmt; Mean RRT start 41.2hrs post-op LOW NOQA=6 Mortality: Early 14/56(25%) died, Late 13/15(87%) died; P=0.0000125 Favors Early
Lundy, 2009 [34] Retrospective cohort US Nov 2005 - Aug 2007 Preexisting renal disease, burn size of less than 40% Non-thermal injury, lithium toxicity Severe Burned patients 57 29 28 Early 27; Late 38 P=0.06 Early: SOFA 13- APACHE2=35; Late: SOFA 13- APACHE2=36 BIOCHEM: AKIN stage 2(+shock)/3; Mean time from admit to RRT = 17 days Mean time from admit to AKIN stage 2(+shock)/3 but not dialyzed = 23 days LOW NOQA=6 28 d mortality: Early 9/29(31%) died, Late 24/28(85%) died; P<0.002; Favors Early
Carl, 2010 [35] Retrospective cohort US 2000-2004 Baseline eGF0R <30ml/min, Age <18 & prisoners Sepsis 147 85 62 Early 52; Late 56 Early: APACHE2=24.8; Late: APACHE2=24.7 BIOCHEM: BUN <100mg/dL + AKIN stage >2; Mean ICU stay prior to RRT =6.3days BUN > 100mg/dL + AKIN stage >2; Mean ICU stay prior to RRT=12.3days HIGH NOQA=7 28 d mortality: Early 44/85(52%) died, Late 42/62(68%); P<0.05 Favors Early
Chou, 2011 [37] Retrospective cohort ‘NSARF’ database Taiwan Jan 2002 - Oct 2009 Age< 18, ICU stay <2days, RRT < 2days Sepsis + AKI 370 192 178 Early 64; Late 66 Early: SOFA 10.8- APACHE2=12.3; Late: SOFA 11.6- APACHE2=14.0 BIOCHEM: RIFLE criteria: RISK or pre-RISK criteria RIFLE criteria: INJURY or FAILURE criteria LOW NOQA=6 Hosp mortality: Early 135/192(71%) died, Late 124/178 (70%) died (P=0.98)
Vats, 2011 [38] Retrospective cohort USA Jan1999 - Feb 2006 Renal transplant, Pre-morbid ESRD on dialysis, RRT<24h, insufficient data Multisystem 230 NR NR All patients mean = 66 NR NR TIME: Time from AKI to RRT < 6 days Time from AKI to RRT≥6d LOW NOQA=5 OR for Late Mortality (>6d) 11.66 (1.26-107.9) P=0.0305, Favors Early
Ji, 2011 [36] Retrospective cohort China Ap 2004 - Mar 2009 Patients readmitted post discharge, Discharged against medical advice, Death <24hrs Post cardiac surgery 58 34 24 Early 64; Late 62 Early: APACHE3=69.3; Late: APACHE3=88.2 p<0.001 TIME: Time from urine output <0.5ml/kg/h to RRT<12h; Mean oliguria to start of RRT 8.4hrs Urine output <0.5ml/kg/h & Time to RRT>12h post oliguria; Mean oliguria to start of RRT 21.5hrs LOW NOQA=6 Hosp mortality: Early 3/34 (9%) died, Late 9/24 (37%); p=0.02 Favors Early
Shiao, 2012 [41] Retrospective cohort ‘NSARF’ database Taiwan Jan 2002 - Apr 2009 Dialysis before surgery, ESRD Surgical 648 436 212 Early 62; Late 66; P=0.009 Early: SOFA 11.4- APACHE2=12.7; Late: SOFA 11.3- APACHE2=12.8 TIME: Time to development of tradtional RRT indications < 3d; Mean time to start of RRT 1.4days Traditional RRT indications AND start of RRT > 3 days; Mean time to start of RRT 18days LOW NOQA=6 Hosp mortality: Early 236/436 (54%) died, Late 143/212 (67%) died; P=0.001 Favors Early
Chon, 2012 [40] Retrospective cohort South Korea Apr 2009 - Oct 2010 Liver cirrhosis, Pre existing chronic Sepsis 55 36 19 Early 63; Late 62 Early: SOFA 13.5- APACHE2= 28.7; Late: SOFA 12- APACHE2=28.3 TIME: Time to RIFLE ‘Injury’/‘Failure’ < 24hrs; Mean time to RRT=12.5hrs Time to RIFLE ‘Injury’/‘Failure’ > 24hrs; Mean time to RRT= 42.2hrs HIGH NOQA=7 28 d mortality: Early 7/36(38%), Late 9/19(47%); P=0.03 Favors Early
Boussekey, 2012 [39] Retrospective cohort France Jan 2008 - Dec 2010 Early trasfer to another unit Multisystem 110 67 43 Early 62; Late 66 Early: SOFA: 11.1- SAPS2=70; Late: SOFA 8.8- SAPS2=57; p=0.002 TIME: Time from RIFLE- ‘Injury’ to RRT < 16hrs; Mean time to RRT=6hrs Time from RIFLE-‘Injury’ to RRT > 16hrs; Mean time to RRT=64hrs LOW NOQA=7 28 d mortality: Early-28/67 (41%), Late- 28/43 (65%); P = 0.0425 Favors Early
Suzuki, 2013 [43] Retrospective cohort Japan Jan 2009 - Feb 2013 <18, RRT for ESRD Sepsis, Cardiogenic Shock 189 52 137 All patients mean = 72 NR All patients SAPS II Mean= 57 BIOCHEM: RIFLE ‘Risk’ RIFLE ‘Injury’ or ‘Failure’ LOW NOQA=6 Early: OR 0.361 (95 % CI 0.17–0.78); P = 0.009, Favors Early
Shum, 2013 [43] Retrospective cohort China Jan 2008 - Jun 2011 Age<18, Chronic dialysis, RRT prior to ICU Sepsis 120 31 89 qEarly 74; Late 73 Early: SOFA 12- APACHE4=119; Late: SOFA 13- APACHE4=133; P=0.011 BIOCHEM: sRIFLE-‘pre- Risk’ or ‘Risk’ criteria; Mean time from ICU admit to RRT =20.7hrs, P=0.056 sRIFLE ‘Injury’ or ‘Failure’ criteria; Mean time from ICU admit to RRT=10.8hrs LOW NOQA=6 28 d mortality: Early-15/31 died (48.4%), Late- 43/89 died (48.3%); P=0.994
Tian, 2014 [46] Retrospective cohort China Nov 2009 - Dec 2011 Age < 12, Chronic renal disease, Terminal illness,0 Pre-admit CRRT, ICU stay < 72hrs Sepsis - AKIN 1 49 23 26 Early 48; Control 54 Early: SOFA 7.6- APACHE2=12.9; Control: SOFA 8.4- APACHE2=15.3 BIOCHEM: AKIN 1 (Cr≥26.4μmol/L or >150- 200% baseline & urine <0.5cc/kg/h for >6h) No RRT (Control): Patients refused CRRT for “personal reasons” LOW NOQA=6 28 d mortality: Early 5/23(22%) died, Control 11/26 (42%) died (NS)
      Sepsis - AKIN 2 52 31 21 Early 54; Control 61 Early: SOFA 9.3- APACHE2=19; Control SOFA 9.6- APACHE2=18.3 AKIN 2 (Cr>200-300% baseline & urine <0.5cc/kg/h for >12h) No RRT (Control): Patients refused CRRT for “personal reasons”   28 d mortality: Early 12/31 (39%) died, Control 14/21 (67%) died; P<0.05 Favors Early
      Sepsis - AKIN 3 59 46 13 Early 50; Control 55 Early SOFA 10- APACHE2=21.8; Control SOFA 11.2- APACHE2=20.5 AKIN 3 (Cr≥354μmol/L or Cr>300% baseline w/urine <0.3cc/kg/h for 24h or anuria >12h) No RRT (Control): Patients refused CRRT for “personal reasons”   28 d mortality: Early 31/46(67%) died, Control 11/13(85%) died; NS
Serpytis, 2014 [45] Retrospective cohort Lithuania 2007-2011 NR Sepsis 85 42 43 All patients mean = 72 NR NR TIME: Time from anuria to RRT < 12hrs Time from anuria to RRT > 12hrs LOW NOQA=5 Mortality: Early 30/42 (71%) died, Late 39/43(91%) died; p=0.028; Favors Early
Gaudry, 2014 [44] Retrospective cohort France Jan 2004 - Nov 2011 Age<18, limitation in medical therapy, death<24hrs, chronic renal insufficiency, RRT prior to ICU, kidney transplant, lithium toxicity, multiple myeloma Sepsis 203 91 112 Early 65; Late 65 Early: SOFA 9- SAPS2=60; Control SOFA 8- SAPS2=55, P<0.01 BIOCHEM: RRT criteria: Cr≥300μmol/L, Urea>25mmol/L, K>6.5mmol/L, pH<7.2, Oliguria, Vol overload, No RRT initiated/Criteria not met for RRT LOW NOQA=5 Hosp Mortality: Early 44/91(48%) died, Control (No RRT) 29/112 (26%) died; P<0.001 Favors no RRT
Retrospective TOTALS       2841 1434 1177       Pooled mortality: Early 714/1434 (50%), Late 732/1177 (62.2%); n=19
  1. LEGEND: AKI Acute kidney injury, AKIN Acute Kidney Injury Network, APACHE Acute Physiology and Chronic Health Evaluation, Cr Creatinine, CRF Chronic renal failure, CRRT Chronic renal replacement therapy, eGFR Estimated glomerular filtration rate, EHV Early High Volume, ELV Early Low Volume, ESRD End stage renal disease, ICU Intensive Care Unit, LLV Late Low Volume, NOQA Newcastle-Ottawa quality assessment, NR Not reported, NSARF National Taiwan University Hospital-Surgical ICU- Acute Renal Failure database, RIFLE Risk, Injury, Failure, Loss and End-stage, RPGN Rapidly progressive glomerularnephritis, SAPS2 Squential Acute physiology Score, SHARF Stuivenberg Hospital Acute Renal Failure Score, SOFA Sequential Organ Failure Assessment, UOP Urine output