====================================================================================

ADDITIONAL FILE 5 SUMMARY ¡ª 20 external studies plus the present study

====================================================================================



Handling of concurrent events

   0/20  ( 0.0%)  used a landmark

   0/20  ( 0.0%)  excluded cases already meeting AKI criteria in the exposure window

                     counts only explicit exclusion of prevalent events, not exclusion of pre-existing CKD or dialysis

   0/20  ( 0.0%)  reported how many events had already occurred by the end of the exposure window



Reporting of the time origins

  15/20  (75.0%)  named a time origin for the exposure

  15/20  (75.0%)  named a time origin for the outcome

  11/20  (55.0%)  named both

   1/20  ( 5.0%)  named neither



When the exposure was measured

   9/20  (45.0%)  exposure measured intraoperative

                     Chen 2025; Fu 2026; Hadipourzadeh 2024; He 2023; Lyu 2023; Peng 2025; Wang 2024; Yan 2023; Yang 2020

  10/20  (50.0%)  exposure measured postoperative

                     Flores-Salinas 2024; Lopez-Delgado 2013; Marianello 2026; Radovic 2019; Tang 2026; Wang 2022; Wang 2025a; Wang 2025b; Yu 2023; Zhang 2015

   1/20  ( 5.0%)  exposure measured preoperative

                     Zhang 2021



Effect estimates

  13/20  (65.0%)  reported a numerical effect estimate for lactate

   7/20  (35.0%)  reported none

                     Flores-Salinas 2024; Fu 2026; Hadipourzadeh 2024; Wang 2025a; Wang 2025b; Yan 2023; Yang 2020

   2/20  (10.0%)  estimate is unadjusted

                     Hadipourzadeh 2024; Radovic 2019



Rating

  10/20  (50.0%)  rated LOW

                     Chen 2025; Fu 2026; Hadipourzadeh 2024; He 2023; Lyu 2023; Peng 2025; Wang 2024; Yan 2023; Yang 2020; Zhang 2021

   0/20  ( 0.0%)  rated MODERATE

   4/20  (20.0%)  rated HIGH

                     Marianello 2026; Tang 2026; Wang 2022; Wang 2025b

   6/20  (30.0%)  rated UNCLEAR

                     Flores-Salinas 2024; Lopez-Delgado 2013; Radovic 2019; Wang 2025a; Yu 2023; Zhang 2015



Among studies whose exposure is measured after surgery

  10 postoperative-exposure studies

  5 named both time origins

  5 did not, so window overlap cannot be assessed

  rated: {'UNCLEAR': 6, 'HIGH': 4}



Cohort sizes

  median 426, range 60 to 11624, total 32717



Outputs: <local path>\analysis\results\manuscript_v3

