1.1 What Each Test Actually Measures

Understanding the difference between these tests starts with understanding what proteins they actually detect — not what the textbook calls them.

The critical insight: The immunoassay for urine albumin is specific — it measures albumin and nothing else. The pyrogallol red total protein assay is broad — it measures whatever binds the dye. These are genuinely different measurements of genuinely different analytes, not just two ways of measuring the same thing.

1.2 The Composition of “Normal” and “Pathologic” Urine Protein

Total normal protein is approximately 80-150 mg/24-h, of which albumin is 10-40 mg. So in health, albumin is roughly 10-30% of total urinary protein — not 50%. The “50% rule” enters only when disease shifts the balance toward filtered plasma proteins.

1.3 The Three Mechanistic Patterns of Proteinuria

flowchart TD
    A[Proteinuria detected] --> B{Mechanistic Pattern}
    B --> C[Glomerular<br/>ALBUMIN DOMINANT<br/>Diabetic nephropathy, FSGS,<br/>membranous, IgA, minimal change,<br/>hypertensive nephrosclerosis]
    B --> D[Tubular<br/>LOW MW PROTEINS DOMINANT<br/>Fanconi, lithium, ATN,<br/>aminoglycoside injury,<br/>Dent disease, Lowe syndrome]
    B --> E[Overflow<br/>MONOCLONAL/LIGHT CHAINS<br/>Multiple myeloma, AL amyloid,<br/>MGUS/MGRS, Waldenstrom,<br/>LCDD/HCDD]
    C --> F[ACR detects<br/>PCR detects<br/>Dipstick detects]
    D --> G[ACR MISSES<br/>most of the load<br/>PCR detects<br/>Dipstick may be negative]
    E --> H[ACR MISSES<br/>almost entirely<br/>PCR detects<br/>Dipstick typically negative<br/>UPEP + sFLC REQUIRED]
    style C fill:#a3e4d7
    style D fill:#ffe066
    style E fill:#f5b7b1
    style F fill:#a3e4d7
    style G fill:#ffe066
    style H fill:#f5b7b1

1.4 The 24-Hour Collection — Gold Standard With Real Problems

The 24-hour urine collection has been the reference standard for decades, but it has well-documented limitations:

This is why the field has largely moved away from 24-hour collection as a routine test. The spot ratio normalizes to creatinine specifically to mimic the 24-hour collection without requiring it — and the math works in most patients because creatinine excretion is relatively stable within a patient day-to-day (5-7% CV per Waikar 2018).

2.1 The Rule Has No Single Origin

The clinical aphorism “urinary protein is approximately 50% albumin” appears in textbooks without a landmark validating paper. This review has searched PubMed extensively for a foundational reference and cannot find one. The closest we come is the older observation that in glomerular disease, albumin typically dominates the urinary protein pool — but the proportion varies from 30% in early minimal change to >80% in advanced nephrotic FSGS to near-zero in cast nephropathy. There is no single “50%” that holds across disease states.

2.2 Sumida 2020 — The Definitive Empirical Answer

According to PubMed, Sumida et al. published the largest individual participant meta-analysis on this question in the Annals of Internal Medicine in 2020 #ref1. The CKD-PC consortium analyzed 919,383 adults across 33 cohorts with same-day measures of ACR and PCR or dipstick protein.

2.3 The Critical Finding — “Below PCR 50, No Consistency”

The direct quote from Sumida 2020: “For PCR values above 50 mg/g, the relationship between PCR and ACR was nearly linear on the log scale, with a shallower slope for values greater than 500 mg/g than for those from 50 to 500 mg/g and relative consistency across cohorts. Below a PCR of 50 mg/g, little consistency in association was seen across cohorts.”

Translation: The conversion between PCR and ACR works reasonably well in patients with clinically meaningful proteinuria (PCR ≥50 mg/g). It does not work in patients with low-grade proteinuria — which is exactly the population where screening decisions are made.

2.4 Diagnostic Performance — Sumida 2020 Numbers

Sensitivity and specificity of PCR thresholds for ACR-based CKD screening and staging:

The A2 staging sensitivity is where dipstick and PCR both fail. Dipstick at trace-to-+ only catches 36% of A2 patients. PCR conversion catches 75%. Both miss a clinically meaningful fraction at the critical moderately-increased-albuminuria threshold.

2.4a Practical Conversion Table — What Each PCR Value Means for ACR, and Why the 50% SWAG Hides the Variability That Actually Matters

The textbook teaches that urinary protein is “approximately 50% albumin.” That single-number rule is a Stone-Walled Approximate Guess — a SWAG — not a validated equation. The Sumida 2020 crude model (Table 2 of the paper, computed here for the clinically common PCR range) gives the population-average empirical answer. But a single median value hides the individual-patient scatter that is what clinicians actually see at the bedside. Both matter, and the contrast between them is where the 50% rule fails hardest.

flowchart TD
    A[uPCR result<br/>Can we convert to ACR?] --> B{uPCR<br/>less than<br/>0.05 g/g<br/>50 mg/g?}
    B -- Yes --> C[STOP<br/>Below Sumida floor<br/>Inconsistent across cohorts<br/>Order dedicated ACR]
    B -- No --> D{Suspected<br/>paraprotein<br/>myeloma MGRS<br/>AL amyloid?}
    D -- Yes --> E[STOP<br/>Light chains missed<br/>by ACR and by conversion<br/>Order UPEP SPEP sFLC<br/>Leung 2019 IKMG]
    D -- No --> F{Suspected<br/>tubular pattern<br/>Fanconi lithium ATN?}
    F -- Yes --> G[STOP<br/>Non-albumin proteins dominate<br/>Conversion underestimates ACR<br/>Order UPCR + dedicated ACR<br/>β2M α1M at academic centers]
    F -- No --> H{Does KDIGO staging<br/>A2 vs A3<br/>matter for this<br/>clinical decision?}
    H -- Yes --> I{Is uPCR<br/>0.5-1.0 g/g<br/>decision window?}
    I -- Yes --> J[ORDER DEDICATED ACR<br/>Sumida 80 percent PI<br/>CROSSES A2 A3 boundary<br/>in this exact window<br/>Cannot stage from PCR alone]
    I -- No --> K[Conversion table is reliable<br/>PCR greater than or equal to 1.25 g/g → A3<br/>PCR less than or equal to 0.25 g/g → A2]
    H -- No --> L[Conversion acceptable<br/>for trend monitoring<br/>or presence of<br/>albuminuria confirmation]
    style A fill:#ffe066
    style C fill:#f5b7b1
    style E fill:#f5b7b1
    style G fill:#f5b7b1
    style J fill:#ffcc99
    style K fill:#a3e4d7
    style L fill:#a3e4d7

2.4b Why ACR Is the Therapy-Guiding Measure — Variability Stacking

The conversion table above answers a staging question. It does not fully answer the therapy-monitoring question, which is subtly different. When you use PCR-to-ACR conversion to track a patient on GDMT — SGLT2i, RAAS blockade, finerenone, endothelin antagonists — you are stacking two sources of uncertainty on top of each other, and the noise compounds.

PCR-as-surrogate-for-ACR has two layers of uncertainty. ACR measured directly has only one.

The within-patient biological variability (Layer 2) is real and unavoidable for both measures — a single UACR can rise or fall +124% / −55% on retest without any true biological change (Waikar 2018; see Section 3.1). That layer does not go away no matter what test you order. But the conversion-scatter layer (Layer 1) is additive and avoidable. Ordering ACR directly removes it.

flowchart TD
    A[CKD patient on GDMT<br/>SGLT2i RAASi finerenone<br/>Need to track response] --> B{Which test<br/>to order?}
    B -- Order PCR<br/>convert to ACR --> C[Layer 1<br/>CONVERSION SCATTER<br/>Sumida 80 percent PI<br/>approximately 28-69 percent of PCR<br/>Methven 2010 IQR 12-64 percent]
    C --> D[Layer 2<br/>BIOLOGICAL VARIABILITY<br/>Waikar 2018 CV approximately 30 percent<br/>RCV +124 / −55 percent<br/>posture exercise hydration fever]
    D --> E[NOISY ACR ESTIMATE<br/>30 percent treatment response<br/>MAY BE MASKED<br/>or spuriously inflated]
    B -- Order ACR<br/>directly --> F[Layer 1<br/>NONE<br/>direct albumin immunoassay]
    F --> G[Layer 2<br/>BIOLOGICAL VARIABILITY<br/>same as PCR pathway<br/>unavoidable for any test]
    G --> H[SHARPER MEASUREMENT<br/>30 percent reduction<br/>DETECTABLE per Heerspink 2019<br/>surrogate endpoint validation]
    E --> I{Intensify GDMT?<br/>Add MRA finerenone<br/>Add ERA<br/>Optimize RAASi SGLT2i}
    H --> I
    style A fill:#ffe066
    style C fill:#ffcc99
    style D fill:#e0e0e0
    style E fill:#f5b7b1
    style F fill:#a3e4d7
    style G fill:#e0e0e0
    style H fill:#a3e4d7
    style I fill:#ffe4b5

2.5 Methven 2010 NDT — The Non-Linearity Confirmed In Nephrology Clinic

Methven et al. confirmed the ACR-PCR non-linearity in a single-center CKD cohort of 6,842 patients #ref2. The direct quote: “The relationship between ACR and TPCR is non-linear. As expected, ACR is almost always less than TPCR. The relationship between ACR and non-albumin protein:creatinine ratio (NAPCR) is poor, with wide scatter, making it difficult to predict TPCR from ACR.”

Key numerical findings:

TPCR outperforms ACR for predicting 24-h protein at clinically meaningful thresholds (AUC 0.967 vs 0.951, P<0.001). The effect is small — both tests are excellent — but the direction is opposite to what the “50% rule” would predict.

The age/sex dependence is striking. To achieve 95% sensitivity for detecting 1 g/day proteinuria, the required TPCR threshold varies: - Man under 49 years: TPCR 65 mg/mmol (approximately 575 mg/g) - Woman over 79 years: TPCR 179 mg/mmol (approximately 1,582 mg/g)

This 2.7-fold variation in threshold by demographics is a direct consequence of the creatinine denominator: elderly women have lower muscle mass and lower creatinine excretion, so the same absolute protein excretion produces a higher PCR. ACR has the same issue but partially compensated by normalization.

3.1 Waikar 2018 AJKD — The Variability Anchor Paper

According to PubMed, Waikar et al. measured short-term within-person variability in 50 clinically stable outpatients with CKD, collecting serial samples across 3 visits within 4 weeks #ref3.

This table is the single most important dataset in this review for answering Andy’s question about spot variability.

What RCV means in practice: If a patient’s random spot UACR is 150 mg/g at visit 1 and 330 mg/g at visit 2, that is a 120% increase — which falls inside the reference change range (+124%). It is NOT statistically significant at P=0.05. You cannot say that patient’s albuminuria has “worsened” without a third measurement or a change that exceeds the RCV.

3.1a Supporting Evidence — Rasaratnam 2024 and Naresh 2013

Two complementary papers extend the Waikar findings and were identified during OpenEvidence verification of the claims in this review:

• Rasaratnam et al. 2024 AJKD (n=826 T2D patients) #ref21 reported a within-person CV of 48.8% for UACR — higher than Waikar’s 30% and consistent with the fact that the T2D population has more metabolic variability affecting albumin excretion. The paper explicitly states that “a repeat UACR measurement can be as low as 0.26 times or as high as 3.78 times the initial value due to random biological variation alone.” The copy in Andy’s DEVONthink Research Articles database is this paper.
• Naresh et al. 2013 AJKD #ref22 quantified reference change values by albuminuria category: for normoalbuminuria, a ±467% change is required for 95% statistical significance; for microalbuminuria, ±170%. These numbers are even more stringent than Waikar’s and confirm that the lower the baseline albuminuria, the larger the percentage change needed to call it a true change.

3.2 Pugliese 2011 RIACE — The Counterweight

According to PubMed, Pugliese et al. examined reproducibility in 4,062 T2D patients from the Italian RIACE multicenter study #ref4. Patients provided 2-3 urine samples over 3-6 months — a longer window than Waikar, closer to a real clinical follow-up interval.

The Pugliese take is more nuanced than Waikar. Yes, within-person CV is approximately 32% (essentially identical to Waikar). BUT for the purpose of categorizing a patient into normo, micro, or macroalbuminuria, a single measurement is 84-95% concordant with the multi-sample geometric mean. For population-scale classification and staging, a single UACR works well enough. The problem is the individual patient whose single value sits near a threshold.

3.3 First-Morning vs Random Spot — Which Is Better?

Both Waikar 2018 and the Hayashi 2013 validation study in Japanese CKD patients (n=159) agree on the answer: first-morning is modestly better than random spot, primarily because it avoids diurnal variation and postural (orthostatic) proteinuria.

From Waikar 2018: - First-morning UACR median CV 32.5% vs random spot UACR 29.7% — actually similar - But first-morning UAC CV 33.2% vs random spot UAC 50.6% — first-morning far better for the unadjusted concentration

The reason first-morning and random UACRs appear similar is that the creatinine denominator absorbs much of the diurnal variation. When you look at albumin concentration without creatinine normalization, the random spot is markedly worse.

KDIGO 2024 recommends first-morning spot UACR as the preferred sample for practical reasons: 1. Standardizes the collection time across patients 2. Avoids postural proteinuria (which resolves overnight in recumbency) 3. Concentrates proteins from overnight urine, improving assay sensitivity 4. Fits into a routine morning clinic workflow

4.1 The Study

According to PubMed, Methven et al. analyzed 5,586 CKD patients attending the Glasgow Royal Infirmary nephrology clinic with contemporaneous ACR and TPCR measurements #ref5. Median follow-up was 3.5 years. During follow-up: - 844 patients (15%) died at median 3.0 years - 468 patients (8%) started renal replacement therapy at median 1.7 years

4.2 The Four Groups

4.3 The Headline Finding

Direct quote from Methven 2011: “Patients with significant proteinuria by TPCR, but not ACR (n=231) had high renal risk, and the highest all-cause mortality (log-rank P<0.001). With multivariate analysis the risk fell below those with significant proteinuria with concordant results by ACR and TPCR but remained considerably higher than those without significant proteinuria.”

Translation: If you only measure ACR in non-diabetic CKD, you miss 231 out of every 1,400 patients with significant proteinuria — approximately 16% — and those missed patients have the highest mortality in the cohort. They are not low-risk false positives. They are high-risk true positives that the ACR test cannot detect.

flowchart TD
    A[CKD Patient<br/>Order ACR only] --> B{ACR result}
    B -- Normal --> C[Reassure<br/>Standard CKD care]
    B -- Elevated --> D[Standard<br/>proteinuric CKD pathway]
    C --> E[MISSED GROUP<br/>16% have significant<br/>TPCR but normal ACR<br/>HIGH mortality<br/>HIGH renal risk]
    style A fill:#ffe066
    style D fill:#a3e4d7
    style E fill:#f5b7b1

4.4 Why the Discordant Group Has Worse Outcomes

The most likely explanation: these are patients with tubulointerstitial disease, paraproteinemia, or monoclonal gammopathy — conditions where non-albumin proteinuria is the dominant signal. These disease states often have: - Heavier total proteinuria than the ACR suggests - Lower eGFR at diagnosis - Higher rates of myeloma cast nephropathy, amyloidosis, or LCDD - Different therapeutic windows (lenalidomide/bortezomib for myeloma, not RAAS blockade)

An ACR-only screening strategy systematically under-diagnoses these patients and delays the correct workup.

5.1 The Foundational Meta-Analyses

Four CKD Prognosis Consortium (CKD-PC) meta-analyses provide the empirical foundation for the KDIGO risk-stratification heat map. Each targets a different population (general, high-risk, CKD-specific, or updated global), which is why all four are cited together:

5.2 Matsushita 2010 Lancet — The Original Heat Map Evidence

According to PubMed, the 2010 CKD-PC general-population meta-analysis established the independent and multiplicative associations of eGFR and UACR with all-cause and cardiovascular mortality #ref6.

Key finding — all-cause mortality HRs (ACR cohorts, n=105,872):

ACR — all-cause mortality HRs:

The relationship between ACR and mortality was linear on the log-log scale without a threshold effect. Unlike eGFR (where risk is flat between 60 and 105), any increase in ACR above the normal range carries graded risk, even at levels well below the traditional 30 mg/g threshold. There is no “safe” level of albuminuria above normal.

The two markers are multiplicatively associated — a patient with eGFR 45 AND ACR 30 has risk that is the product (not the sum) of each individual elevation. This is the empirical basis for the KDIGO heat map.

5.3 Grams 2023 JAMA — The Current Update

According to PubMed, Grams et al. published the most current CKD-PC synthesis in 2023 with 27,503,140 individuals across 114 cohorts #ref7. The paper evaluated 10 adverse outcomes: kidney failure requiring replacement therapy, all-cause mortality, cardiovascular mortality, acute kidney injury, any hospitalization, coronary heart disease, stroke, heart failure, atrial fibrillation, and peripheral artery disease.

Key findings: - Lower eGFR (creatinine-based) and lower eGFR (cystatin C + creatinine) were independently associated with all 10 adverse outcomes - Higher UACR was independently associated with all 10 adverse outcomes - Even in patients with UACR <10 mg/g — below the traditional normal threshold — an eGFR of 45-59 was associated with higher hospitalization rates (adjusted HR 1.3, 161 vs 79 events per 1000 person-years, excess absolute risk 22 events/1000 person-years) - The relationships held across the full range of CKD severity

The Grams 2023 update confirms: the CKD-PC heat map is robust across populations, across outcomes, and across measurement methods (creatinine vs cystatin C for eGFR; UACR for albuminuria). It is the single best prognostic framework nephrology has.

5.4 The KDIGO Heat Map — What the Evidence Looks Like in Practice

flowchart TD
    A[Patient with CKD] --> B[Measure eGFR]
    A --> C[Measure UACR]
    B --> D{eGFR Category}
    C --> E{ACR Category}
    D --> F[G1 ≥90<br/>G2 60-89<br/>G3a 45-59<br/>G3b 30-44<br/>G4 15-29<br/>G5 <15]
    E --> G[A1 <30 mg/g<br/>A2 30-299 mg/g<br/>A3 ≥300 mg/g]
    F --> H[Risk Stratification<br/>via CKD-PC heat map]
    G --> H
    H --> I[Green: low risk<br/>Yellow: moderate<br/>Orange: high<br/>Red: very high]
    I --> J[Drives frequency of monitoring<br/>nephrology referral timing<br/>SGLT2i indication<br/>finerenone indication<br/>BP targets]
    style A fill:#ffe066
    style H fill:#a3e4d7
    style I fill:#ffe4b5
    style J fill:#90ee90

6.1 The CKD-PC Change-in-Albuminuria Meta-Analysis

According to PubMed, Coresh, Heerspink, and the CKD-PC consortium published the landmark meta-analysis validating change in albuminuria as a surrogate for ESKD in Lancet Diabetes Endocrinology 2019 #ref8. The design:

6.2 The Key Finding

A 30% decrease in ACR during a 2-year baseline period was associated with an adjusted hazard ratio of 0.83 (95% CI 0.74-0.94) for subsequent ESKD. After adjusting for regression dilution, the HR became 0.78 (0.66-0.92). The effect was consistent across cohorts and subgroups stratified by eGFR, diabetes status, and sex.

In absolute terms: In patients with baseline ACR ≥300 mg/g (stage A3), a 30% reduction in ACR over 2 years was estimated to produce more than a 1% absolute reduction in 10-year ESKD risk — even at early CKD stages (eGFR >60).

ARR and NNT inference: With a baseline 10-year ESKD risk of, say, 10% in a high-albuminuria CKD3 patient, a 30% ACR reduction produces: - RRR approximately 22% (1 − 0.78) - ARR approximately 2.2% (22% × 10%) - NNT approximately 45 over 10 years to prevent one ESKD event

These numbers are the quantitative basis for the FDA/EMA acceptance of albuminuria change as a surrogate.

6.2a Heerspink 2025 Nature Medicine — The CURRENT Definitive Surrogate Evidence

According to PubMed, Heerspink, Inker, and colleagues published the definitive current trial-level meta-analysis on albuminuria as a surrogate endpoint in Nature Medicine in November 2025 #ref24. This is the most recent and largest evidence base for the 30% UACR reduction threshold and supersedes the 2019 Heerspink paper with nearly three times the participants.

The headline findings:

• Pooled average UACR reduction 25% from baseline in active-treatment arms (GMR 0.75, 95% CI 0.71-0.79)
• Active interventions reduced clinical endpoint by 24% (HR 0.76, 95% CI 0.71-0.82)
• Each 30% reduction in geometric mean UACR → 19% lower hazard for clinical endpoint (95% BCI 5-30%)
• R² = 0.66 (95% BCI 0.06-0.98) — substantially stronger than the 2019 paper’s R² 0.47
• Intercept −0.05 (95% BCI −0.22 to 0.11) — crossing zero, meaning NO treatment effect on clinical endpoint when there is NO UACR reduction (biological coherence)
• Posterior probability that slope >0: 99.5%
• For 30% UACR reduction, posterior probability of beneficial clinical effect: >99%

6.2b Heerspink 2019 Lancet Diabetes Endocrinol — The Historical Context

According to PubMed, the companion trial-level paper to the Coresh 2019 observational meta-analysis was published in the same issue of Lancet Diabetes Endocrinology #ref23. It is now superseded by the 2025 Nature Medicine update but remains the foundational reference for how the surrogate framework was developed.

The headline finding: “Each 30% decrease in geometric mean albuminuria by the treatment relative to the control was associated with an average 27% lower hazard for the clinical endpoint (95% BCI 5-45%; median R² 0.47, 95% BCI 0.02-0.96).”

Predicted performance for future trials: “Treatments that decrease the geometric mean albuminuria to 0.7 (ie, 30% decrease in albuminuria) relative to the control will provide an average hazard ratio (HR) for the clinical endpoint of 0.68, and 95% of sufficiently large studies would have HRs between 0.47 and 0.95.”

The key conditional: The surrogate association strengthens dramatically when restricted to patients with baseline ACR >30 mg/g (R² improves from 0.47 to 0.72), which matches the Levey 2020 NKF/FDA/EMA workshop findings. In patients with low baseline albuminuria, the surrogate relationship is uncertain. This is a critical caveat — UACR reduction is a valid surrogate in already-proteinuric patients, not in normoalbuminuric patients whose trials are powered on absolute UACR rise.

What trial sample size means for the threshold: The Heerspink/Inker analysis also determined that for participants with baseline ACR >30 mg/g, a 20% decrease in ACR would be required in studies of infinite sample size, 21% in large trials, and 27% in modest-sized trials to demonstrate clinical benefit. These numbers justify the 30% reduction threshold used as an operational target across the SGLT2i, GLP-1 RA, finerenone, and endothelin antagonist trial programs.

6.2c Heerspink 2023 JASN — Joint UACR + GFR Slope Surrogates

According to PubMed, Heerspink, Inker, and the CKD-EPI Clinical Trials consortium published the joint surrogate framework in JASN 2023 #ref25. The idea: combining treatment effects on UACR change AND chronic GFR slope improves prediction of clinical endpoints better than either surrogate alone.

The headline finding: - Combined treatment effects on UACR change + chronic GFR slope strongly predicted clinical endpoint effects - Coefficient for GFR slope: −0.41 per 1 mL/min/1.73 m²/year (95% BCI −0.64 to −0.17) - Coefficient for UACR change: −0.06 (95% BCI −0.90 to 0.77, wider and weaker alone) - In small-sample short-follow-up Phase 2 trials (approximately 60 patients per arm, 1 year), UACR change is the dominant predictor. GFR slope importance increases with larger sample sizes and longer follow-up.

The operational take: For Phase 2 CKD drug development, UACR change is the more actionable surrogate in smaller/shorter trials, while GFR slope becomes more informative as the trial grows. This is why most Phase 2 ASi, ERA, and finerenone-adjacent trials in recent years have reported BOTH UACR change (early, at 6-12 weeks) AND eGFR slope (at 6+ months) as co-primary surrogates.

6.2d How This Plays Out in Current Therapy Trials

The practical validation of the 30% UACR reduction target comes from the recent aldosterone synthase inhibitor program. In the aldosterone synthase inhibition trial portfolio (see companion baxdrostat aldosterone synthase inhibitor medical review Section 4 for the full trial-by-trial breakdown), 51% of participants on monotherapy and 70% of those also receiving empagliflozin achieved a 30% or higher UACR reduction from baseline — confirming that the surrogate target is clinically achievable with modern therapy combinations. This is the “additive UACR reduction” framework that drives the rationale for combining SGLT2i with MRAs, ASis, and endothelin antagonists.

6.3 Levey 2020 AJKD — The NKF/FDA/EMA Workshop

According to PubMed, the NKF/FDA/EMA joint scientific workshop published its consensus in AJKD 2020 #ref9. The workshop evaluated changes in albuminuria and eGFR slope as candidate surrogate endpoints through three parallel efforts: meta-analyses of observational cohorts, meta-analyses of clinical trials, and simulation modeling.

Key consensus findings: - A UACR reduction of 30% is associated with an HR of approximately 0.7 for clinical kidney outcomes — consistent in cohorts and trials - eGFR slope reduction of 0.5 to 1.0 mL/min/1.73 m²/year is associated with similar HR - Posterior median R² of treatment effect on change in UACR vs clinical outcome: 0.47 (95% BCI 0.02-0.96), rising to 0.72 when restricted to baseline UACR >30 mg/g - eGFR slope has stronger surrogacy properties than change in UACR (posterior R² 0.97 for 3-year slope)

The workshop’s regulatory position: Both early change in albuminuria and GFR slope fulfill criteria for surrogate use under specific conditions — stronger support for eGFR slope than for albuminuria change. Albuminuria change is accepted when baseline UACR is ≥30 mg/g and the treatment effect pathway is plausible.

6.4 Carrero 2017 Kidney International — The Concordant Finding

According to PubMed, Carrero, Coresh, and colleagues published a complementary analysis in Kidney International 2017 showing that albuminuria changes are associated with subsequent risk of end-stage renal disease and mortality in population-based cohorts #ref10. This paper extends the surrogate evidence to the broader patient population and supports the generalizability of the CKD-PC findings.

6.5 Heerspink-Gansevoort 2015 — The Pro Case

According to PubMed, Lambers Heerspink and Gansevoort’s CJASN 2015 “Pro View” article argued that albuminuria is an appropriate therapeutic target in CKD #ref11. The mechanistic case: - Elevated albuminuria causes tubulointerstitial damage through activation of proinflammatory mediators - Reduction in albuminuria during treatment correlates with long-term renal protection - Residual albuminuria in treated patients remains the strongest prognostic marker - The initial magnitude of UACR reduction predicts long-term ESKD risk — supporting the “responder analysis” trial designs seen in SONAR, FIDELIO, and others

7.1 Light Chain / Paraproteinemia / Myeloma — 1-Paragraph Summary

Light chains are neither albumin nor standard globulins and are systematically missed by UACR. Multiple myeloma, AL amyloidosis, light chain deposition disease (LCDD), and monoclonal gammopathy of renal significance (MGRS) can all present with kidney injury and minimal UACR despite substantial proteinuria by UPCR. The Leung 2019 IKMG consensus recommends serum and urine protein electrophoresis (SPEP/UPEP), immunofixation, and serum free light chain (sFLC) assays as the complete workup when paraproteinemia is suspected — not as an “add-on” to UACR but as independent, necessary tests #ref12. The single most actionable rule: if a patient has unexplained proteinuria by UPCR or dipstick but a normal UACR, order UPEP and sFLC before you do anything else. Cast nephropathy, AL amyloid, and MGRS are missed diagnoses with substantial mortality when treated as “idiopathic nephrotic syndrome.”

7.2 Pre-Eclampsia — 1-Paragraph Summary

The 24-hour urine collection remains the gold standard for diagnosis of pre-eclampsia proteinuria, with ≥300 mg/24-h defining significant proteinuria. ACOG and SMFM accept spot PCR ≥0.3 (mg/mg) as a diagnostic equivalent in the setting of new-onset hypertension after 20 weeks gestation when 24-h collection is impractical. Dipstick ≥1+ on two occasions is the minimum acceptable substitute. UACR is not the preferred test in pregnancy because tubular and Tamm-Horsfall proteins contribute meaningfully in gravid patients, and the evidence base for diagnostic thresholds rests on total protein, not albumin. Practical rule: in suspected pre-eclampsia, if you have time for a 24-h collection, get one; if you do not, use spot PCR with a 0.3 cutoff, not UACR.

7.3 Diabetic Nephropathy

Diabetic nephropathy is the textbook “ACR works” disease. The injury pattern is dominantly glomerular with albumin as the predominant protein. UACR correlates tightly with 24-h albumin excretion. The Mogensen microalbuminuria paradigm — progression from normoalbuminuria through microalbuminuria (30-299 mg/g) to overt nephropathy (≥300 mg/g) — was built on albumin measurements specifically.

ADA recommends annual UACR screening for all adults with T1DM ≥5 years and all adults with T2DM at diagnosis and annually thereafter. First-morning spot UACR is the recommended sample.

7.4 Glomerular Diseases — Minimal Change, FSGS, Membranous, IgA

All four are glomerular, albumin-dominant. UACR works well for screening and monitoring. The exception: during nephrotic syndrome (UACR >2200 mg/g roughly corresponding to 3+ g/day), the absolute gram-per-day matters for diagnosis (nephrotic syndrome requires ≥3.5 g/day per most definitions) and clinical thresholds (some trials enrolled by 24-h protein, not ACR). In the nephrotic range, many centers still collect a baseline 24-h for clinical trial eligibility and treatment response documentation.

7.5 Tubular Diseases — Fanconi, ATN, Lithium

Tubular injury produces a non-albumin proteinuria dominated by low-molecular-weight markers (β2-microglobulin, α1-microglobulin, retinol-binding protein). UACR dramatically underestimates total proteinuria in these patients. UPCR is the screening test of choice. Specific tubular markers (β2M, α1M) are available at academic centers for confirmation but rarely change management.

Lithium nephrotoxicity is a particularly important case: patients on long-term lithium develop interstitial nephritis with modest proteinuria and concentrating defects. UACR may be normal or mildly elevated while UPCR shows clinically significant protein. Always get UPCR on lithium patients at baseline and annually.

7.6 Hypertensive Nephrosclerosis

Mixed glomerular and tubular injury pattern. UACR is the recommended screening test per KDIGO and ADA, but UPCR adds information at any given level of ACR — the Methven 16% group likely includes many hypertensive nephrosclerosis patients who have elevated TPCR beyond what their ACR would predict.

7.7 Cardiorenal Syndrome

Proteinuria in cardiorenal syndrome is predominantly hemodynamic (elevated venous pressure → glomerular congestion) rather than intrinsic glomerular injury. UACR elevation is common and should not automatically trigger a primary glomerular disease workup if the clinical picture fits cardiorenal physiology. The key distinction: albuminuria in decompensated HF often improves with decongestion alone, before any RAAS-blockade dose change. Kidney biopsy is not indicated in the typical cardiorenal HF-CKD patient.

8.1 What KDIGO 2024 Recommends

8.2 What Changed From KDIGO 2012 → 2024

• Greater emphasis on cystatin C for GFR estimation in patients with non-standard muscle mass
• Risk-based approach to kidney failure prediction (using the 4-variable KFRE) now explicitly recommended over eGFR-alone categorization
• SGLT2i are now standard therapy for patients with CKD and albuminuria — the evidence base didn’t exist in 2012
• Point-of-care testing for remote areas acknowledged as a reasonable alternative
• Statin therapy for adults >50 years with CKD reinforced

KDIGO 2024 is compatible with the Methven 2011 critique (PCR adds information ACR misses) but defaults to ACR for practical standardization reasons. The two positions are not contradictory — they address different questions.

8.3 The Madero 2025 Annals Synopsis

According to PubMed, Madero et al. published a concise synopsis of KDIGO 2024 in Annals of Internal Medicine 2025 for the internal medicine audience #ref15. This is the version to hand to PA students and internists who need to use the guideline without reading the full 200+ page document.

9.1 The Decision Algorithm

flowchart TD
    A[Clinical Question:<br/>Does this patient have<br/>kidney disease?] --> B{Known CKD<br/>or Screening?}
    B -- Screening --> C[Order first-morning UACR<br/>as initial test<br/>PLUS dipstick on reflex urinalysis]
    B -- Known CKD --> D[Follow established<br/>monitoring pattern]
    C --> E{ACR result}
    E -- Normal<br/>and dipstick neg --> F[Reassure<br/>Rescreen per guidelines<br/>Annual if DM/HTN]
    E -- Normal<br/>but dipstick positive --> G[ORDER UPCR<br/>Suspect non-albumin<br/>proteinuria]
    E -- Elevated --> H[CONFIRM with<br/>second sample<br/>in 2-4 weeks]
    H --> I{Still elevated?}
    I -- Yes --> J[Stage per KDIGO heat map<br/>Proceed with workup<br/>Initiate RAAS blockade + SGLT2i]
    I -- No --> K[Transient elevation<br/>Monitor, investigate triggers:<br/>exercise, UTI, fever, posture]
    G --> L[UPCR elevated?]
    L -- Yes --> M[Check disease pattern:<br/>Tubular markers<br/>SPEP/UPEP/sFLC<br/>24-h collection<br/>Consider biopsy]
    L -- No --> N[Dipstick false positive<br/>Hematuria, alkaline pH,<br/>contrast, chlorhexidine]
    style C fill:#ffe066
    style G fill:#ffe066
    style H fill:#ffe066
    style J fill:#a3e4d7
    style M fill:#f5b7b1
    style N fill:#e0e0e0

9.2 When to Order What — The Cheat Sheet

9.3 Interpreting Discordant Results

When UACR and UPCR disagree, calculate the urine albumin-to-total-protein ratio (uA/uP):

This ratio comes directly from the Methven 2012 NDT letter #ref16 and is clinically actionable at the bedside. It requires no additional tests — just a same-sample UACR and UPCR.

9.4 The Confirmation Rule

Never act on a single abnormal UACR or UPCR without confirmation. The biological variability literature (Waikar 2018, Pugliese 2011) unanimously supports repeat measurement before: - Starting or escalating RAAS blockade - Starting SGLT2i or finerenone in a borderline patient - Documenting CKD progression - Referring to nephrology - Counseling a patient about prognosis

KDIGO 2024 recommends two to three separate measurements to establish baseline albuminuria status. The cost of getting this wrong (mislabeling a patient as having CKD, or missing a real diagnosis) far exceeds the cost of a repeat spot ACR.