Renal Blood Flow, Glomerular Filtration, and Renal Autoregulation

Renal blood flow supplies a vascular system that must filter plasma while supporting the tissue that modifies the filtrate. Flow, filtration, and urine output are therefore connected without being interchangeable. This chapter explains how the arrangement of renal vessels, glomerular filtering forces, and local feedback allow those quantities to move differently. It also develops the limits of interpreting arterial pressure or creatinine as a direct measure of the underlying renal circulation.

43.1 Two Capillary Beds Give the Renal Circulation Its Special Behavior

Blood reaches the glomerular capillaries through an afferent arteriole and leaves through an efferent arteriole. Thus the filtering capillary bed lies between two resistance vessels. This arrangement permits changes in upstream and downstream resistance to alter glomerular pressure differently, even when both changes reduce total flow.

The efferent circulation then supplies a second capillary network. Cortical nephrons drain predominantly toward peritubular capillaries. Efferent vessels of juxtamedullary nephrons contribute to the vasa recta, which enter the medulla. These vascular pathways support tubular exchange and regional tissue needs. Total renal arterial flow does not specify how much blood traverses each pathway or how evenly capillaries are perfused.

Filtration moves water and sufficiently permeable solutes from glomerular plasma into Bowman’s space. Red cells and most plasma protein remain in the circulation. The filtration barrier includes fenestrated endothelium, basement membrane, and podocyte filtration slits. Filtration volume and barrier selectivity are related but distinct properties: the presence of protein in urine does not directly quantify the volume filtered.

The filtrate subsequently enters the tubule, where most filtered water and many solutes return to blood. Other substances are secreted from blood into tubular fluid. Final urine therefore represents the result of filtration, reabsorption, and secretion rather than a sample of the original filtrate.

This architecture explains several clinically important dissociations. Glomerular pressure can be maintained while renal plasma flow falls. Urine volume can decline because water reabsorption rises despite continued filtration. Conversely, filtration can fall while total renal flow is preserved. Each pattern requires a different explanation of vascular and tubular behavior.

Whole-kidney measurements also sum the behavior of many nephrons. A preserved total GFR can conceal different single-nephron filtration rates, and an injured kidney need not lose function uniformly across its cortex and medulla.

43.2 Blood Flow, Plasma Flow, and Filtration Fraction Need Different Denominators

Renal blood flow, RBF, is whole blood passing through the kidneys per unit time. Renal plasma flow, RPF, excludes the red-cell volume. Hematocrit is the fraction of whole-blood volume occupied by red cells, so the plasma fraction is one minus hematocrit.

Consider illustrative combined flow to both kidneys of 1,000 mL/min and hematocrit 0.40. The plasma fraction is 0.60, giving RPF of 1,000 × 0.60 = 600 mL/min. This arithmetic assumes the stated hematocrit adequately represents blood entering the kidneys. The values illustrate definitions rather than a required normal flow.

Glomerular filtration rate, GFR, is the volume of ultrafiltrate produced per unit time. Filtration fraction, FF, divides GFR by incoming plasma flow, using compatible units.

If GFR is 120 mL/min in the example, FF is 120/600 = 0.20, or 20%. Dividing by whole-blood flow instead would give 12%, which answers a different question. FF describes the fraction of incoming plasma volume filtered; it is not the fraction excreted as urine or an index that independently establishes oxygen adequacy.

Measurement methods must also be distinguished. Clearance expresses a urinary excretion rate divided by plasma concentration. A freely filtered marker without secretion or reabsorption can estimate GFR under appropriate sampling conditions. Para-aminohippurate, or PAH, undergoes filtration and tubular secretion, allowing its clearance to estimate effective plasma flow when extraction is high. The foundational clearance studies established this approach. Incomplete or changing extraction requires correction before effective flow is equated with total RPF. 1

Doppler velocity is another quantity: estimating volume flow also requires vessel area and suitable sampling. A velocity waveform or resistive index is not a direct measurement of total RBF, and none of these flow methods directly measures GFR.

43.3 Glomerular Filtration Depends on Opposing Pressures and Kf

Glomerular capillary hydrostatic pressure favors filtration into Bowman’s space. Hydrostatic pressure in Bowman’s space opposes that movement. Plasma proteins exert an oncotic pressure that retains water within the capillary. Oncotic pressure in normal low-protein filtrate is small enough to approximate as zero in a simplified calculation.

The ultrafiltration coefficient, Kf, combines the available filtering surface area and its hydraulic permeability. Multiplying Kf by the effective net filtering pressure estimates filtration. The symbols below denote glomerular capillary hydrostatic pressure, P_GC; Bowman’s-space hydrostatic pressure, P_BS; and the corresponding oncotic pressures, π_GC and π_BS.

This compact form requires a representative effective pressure across the filtering surface. Local conditions change along a capillary. As relatively protein-free fluid leaves, remaining plasma protein becomes more concentrated and capillary oncotic pressure rises. The net driving pressure consequently tends to decline toward the efferent end. Hydrostatic pressure can also vary, so one fixed numerical difference cannot describe every location.

Brenner, Troy, and Daugharty directly measured glomerular pressures in surface glomeruli of a specialized rat strain, relating the hydrostatic and oncotic forces to single-nephron filtration. These experiments established the value of measuring the components rather than inferring them solely from systemic pressure. Their rat pressures and coefficients are not human bedside reference values. 2

For an illustrative whole-kidney model, let the effective hydrostatic difference be 40 mm Hg and the opposing oncotic difference 30 mm Hg. Net pressure is 10 mm Hg. With assumed Kf of 12 mL/min/mm Hg, filtration is 12 × 10 = 120 mL/min. If net pressure falls to 5 while Kf remains fixed, modeled filtration becomes 60 mL/min. In a living kidney, pressure, plasma flow, and Kf can change together; the example isolates one dependency.

43.4 Afferent and Efferent Tone Can Separate Flow From Filtration

Constriction before the glomerulus increases the pressure loss between the renal artery and glomerular capillaries. With other conditions held constant, afferent constriction reduces RBF, lowers glomerular hydrostatic pressure, and reduces filtration. Afferent dilation tends to increase flow and transmit more arterial pressure to the glomerulus.

Efferent constriction has a different initial effect. Narrowing the outlet can raise pressure within the glomerular capillary while reducing plasma flow through it. Moderate efferent constriction can therefore preserve or increase filtration despite lower RPF. If filtration is relatively maintained while plasma flow falls, FF rises.

This effect has limits. With more severe flow restriction, the plasma traversing the glomerulus loses a larger fraction of its water, so its protein concentration and opposing oncotic pressure rise more rapidly. Reduced plasma delivery can then outweigh the increased hydrostatic pressure, and GFR falls. A stimulus may also constrict the afferent circulation or reduce Kf, further changing the result.

Angiotensin II illustrates why the outlet-resistance model is useful but incomplete. Its effects depend on concentration, volume state, baseline vascular tone, and local mediators. Blantz and colleagues measured changes in plasma flow, capillary pressure, and ultrafiltration coefficient during angiotensin II exposure in rats. A change in filtering capacity contributed alongside vascular effects; the experiment does not support predicting GFR from efferent tone alone. 3

Efferent dilation can likewise increase renal flow while lowering glomerular pressure and GFR. Thus “renal vasodilation” is not synonymous with improved filtration, just as constriction does not always produce identical proportional falls in flow and GFR.

These are mechanistic comparisons under stated assumptions. A clinical vasoactive intervention also changes systemic pressure, cardiac output, venous pressure, and other renal signals. Its net renal effect cannot be assigned from a single arteriolar label or inferred from a dose without observing the wider response.

43.5 The Myogenic Response Adjusts Resistance When Pressure Changes

If arterial pressure rises while resistance remains unchanged, renal flow and glomerular pressure tend to rise. The myogenic response opposes that transmission. Increased vessel-wall stretch activates smooth-muscle signaling, promotes membrane depolarization and calcium entry, and causes preglomerular constriction. When pressure falls, reduced activation permits dilation.

The adjustment is intrinsic to the vascular wall; it does not require a new instruction from the sympathetic nervous system. It helps stabilize flow and limits transmission of pressure fluctuations to delicate glomerular capillaries. Afferent arterioles make an important contribution, although the preglomerular vascular network is not a single uniform tube.

Takenaka and colleagues studied directly visible afferent arterioles in isolated perfused hydronephrotic rat kidneys. Their experiments supported roles for mechanosensitive signaling, depolarization, and calcium entry in pressure-induced constriction. The preparation separated this response from tubuloglomerular feedback. Pharmacological interference in that model does not establish one exclusive molecular pressure sensor in intact human kidneys. 4

Autoregulation is an active response over time, not instant cancellation of every pressure change. A sudden pressure step can initially alter flow before resistance adjusts. Oscillating pressure and sustained pressure can therefore expose different behavior. The speed and completeness of the response depend on the tissue state and interacting control mechanisms.

Dilation also has a physical limit. Once the relevant vessels are maximally relaxed, a further pressure fall cannot be compensated by the same mechanism. Flow then becomes more pressure dependent. At the high-pressure end, limited protective constriction permits greater pressure transmission.

An arterial pressure that looks acceptable at the monitor does not prove that a particular kidney remains within its effective autoregulatory range. Conversely, a modest pressure change need not produce an equal proportional flow change when regulatory reserve is intact.

43.6 Tubuloglomerular Feedback Connects Tubular Delivery to Afferent Tone

The macula densa is a specialized epithelial region where the returning nephron tubule lies near its own glomerular vascular pole. It senses luminal salt transport rather than measuring arterial pressure directly. This arrangement closes a feedback loop between filtration, upstream tubular handling, and the resistance controlling filtration.

When NaCl delivery and uptake at the macula densa increase, the response tends to constrict the afferent arteriole and reduce single-nephron filtration. This can oppose an initial increase in filtered delivery. Reduced salt uptake diminishes this constrictor signal, permitting filtration to be supported when conditions allow.

Apical NKCC2 transports sodium, potassium, and chloride into macula-densa cells and is central to salt sensing. Experiments demonstrated NaCl-dependent ATP release at the basolateral membrane, connecting luminal transport to a signal outside the cell. 5 Extracellular nucleotide breakdown generates adenosine. The enzyme CD73 catalyzes the final conversion from AMP to adenosine, and feedback was impaired in CD73-deficient mice. 6 Adenosine acting on A₁ receptors near the vascular pole contributes to afferent constriction. Sun and colleagues found loss of the normal feedback response in mice lacking A₁ receptors, providing strong mechanistic evidence in that genetic model. 7

The sensor also has molecular specialization. Oppermann and colleagues selectively deleted the B isoform of NKCC2 in mice. Feedback sensitivity shifted, while other isoforms allowed a substantial response to remain. This supports transporter-dependent sensing without implying that one isoform accounts for all feedback or that a mouse response curve supplies human bedside thresholds. 8

Tubular delivery depends on more than GFR. Changes in proximal reabsorption, tubular flow, salt concentration, and transport inhibition alter what reaches the macula densa. A drug that blocks NKCC2 can therefore interrupt the sensing pathway even while increasing downstream salt delivery. Detailed diuretic effects are developed later.

Local nitric oxide and prostaglandin signals modulate the response and connect low salt uptake to renin release. Feedback strength can change with the physiological state. It is a regulated system that helps match filtration to tubular processing, not a fixed reflex that always restores one GFR.

43.7 Autoregulation Has a Conditional Range Rather Than One Clinical MAP Target

Myogenic control and tubuloglomerular feedback interact. The vascular response reacts to pressure, while the tubular loop responds to altered salt delivery after filtration and transport. Hashimoto and colleagues found less complete stabilization of flow and GFR after pressure reduction in A₁-receptor-deficient mice. This supports an integrated contribution from feedback rather than assigning all autoregulation to the vessel wall. 9

The observed pressure–flow relationship depends on species, anesthesia, volume state, neurohormonal tone, vascular disease, and the duration and method of measurement. A plateau in an experimental graph describes relatively stable flow under those conditions. It does not establish an invariant human MAP interval within which kidney function is protected.

Chronic vascular disease can alter regulatory behavior and the pressure transmitted to glomeruli. Acute inflammation, ischemia, and medications can change it again. A history of hypertension therefore raises a physiological question about the individual response; it does not locate that patient’s lower autoregulatory limit on the monitor.

Clinical pressure-target trials test the consequence of the entire strategy required to produce a pressure. In SEPSISPAM, a higher MAP strategy did not improve overall survival. A chronic-hypertension subgroup required less renal replacement therapy, but this was not accompanied by a demonstrated mortality benefit. 10

The 2025 OPTPRESS trial adds an important qualification. Among older patients with septic shock in Japan, targeting MAP 80–85 rather than 65–70 mm Hg increased 90-day mortality, 39.3% versus 28.6%. These were trial comparison ranges, not universal physiological thresholds. The result argues against turning age or hypertension alone into an instruction to pursue the higher strategy. 11

Neither study directly measured each patient’s autoregulatory curve. A plausible renal-pressure benefit must be judged with systemic effects, treatment exposure, and patient outcomes. Preserving the distinction between mechanism and strategy prevents a textbook plateau from becoming an unsupported clinical target.

43.8 Neural, Hormonal, and Endothelial Signals Modify the Local Response

Autoregulation operates within a wider control system. Sympathetic activation can constrict renal vessels, stimulate renin release, and increase tubular sodium retention. These responses help defend the systemic circulation during stress but can reduce renal flow when activation is strong. The kidney is regulating extracellular volume while also participating in the whole-body response to threatened arterial pressure.

Renin release responds to reduced afferent vascular pressure, sympathetic signaling, and macula-densa conditions. The resulting angiotensin II influences systemic and intrarenal resistance as well as tubular transport. Its filtration effect depends on the balance among afferent resistance, efferent resistance, plasma flow, and filtering capacity described earlier.

Endothelial nitric oxide and locally produced prostaglandins provide vasodilator influences that can counter constrictor signals. They also interact with tubular feedback. Their functional importance changes with the starting state: removing a vasodilator influence during salt depletion or strong neurohormonal activation can produce a different response from removing it under unstressed conditions.

Rossat and colleagues studied cyclooxygenase inhibition in salt-depleted human volunteers. Their observations included transient renal hemodynamic changes and sodium retention, demonstrating that selective cyclooxygenase-2 inhibition did not spare all renal effects in that setting. This is evidence about a physiological dependency in a defined population, not a prediction of identical effects in every critically ill patient. 12

Drug effects can be understood through these mechanisms. Blocking angiotensin signaling can reduce support of glomerular pressure when filtration depends on efferent tone. Inhibiting prostaglandin synthesis can remove a protective vasodilator influence. Neither explanation establishes whether a medicine is appropriate for an individual patient or whether every creatinine change represents structural injury.

The net response also includes changes in arterial pressure, cardiac output, venous conditions, and volume. Mechanistic knowledge makes a change interpretable; it does not replace assessment of the indication, the full clinical state, and treatment outcomes.

43.9 Downstream Pressure Can Limit the Kidney Despite Adequate Arterial Pressure

Blood flow requires a pressure difference across the renal circulation. In a simple open vascular pathway, increasing renal venous pressure reduces that difference if arterial pressure is unchanged. The downstream environment, however, includes more than the pressure measured in a central vein.

Renal venous pressure depends on the venous connection to the heart and on local conditions along the drainage pathway. Abdominal pressure, venous compression, posture, and respiratory pressure transmission can change that relationship. CVP can therefore provide information about the systemic venous environment without being identical to renal venous pressure in every patient.

The kidney also occupies a constrained anatomical space. Venous congestion or tissue edema can raise interstitial pressure, influence small-vessel caliber, and alter tubular and lymphatic drainage. When surrounding pressure becomes important, a collapsible vascular segment can behave differently from a rigid resistance with a freely transmitted downstream pressure.

Raised pressure in Bowman’s space directly opposes filtration. Obstruction of urinary drainage is one route to that back pressure. Interstitial, tubular, and venous pressures interact, but they are not interchangeable measurements and need not rise by the same amount.

Harman and colleagues raised intra-abdominal pressure experimentally in dogs and found substantial reductions in renal flow and filtration. Restoring cardiac output did not restore renal function to its starting state, supporting an important local mechanical contribution. The experimental pressure levels and magnitude of impairment should not be converted into universal human thresholds. 13

Subtracting CVP from MAP can summarize one clinically relevant pressure separation. It is not an exact measurement of the complete renal driving gradient, especially when abdominal or interstitial pressure and vascular compression are important. Nor does improving the subtraction guarantee that filtration has improved, because intrarenal resistance, Kf, and tubular back pressure can still change.

These mechanisms introduce the downstream constraint developed further in Chapter 48. The immediate lesson is to assess inflow and drainage together rather than assigning every filtration decline to insufficient arterial pressure.

43.10 Sepsis Can Dissociate Total Renal Flow From Filtration

Reduced renal blood flow can cause renal dysfunction, but septic AKI does not require a uniform fall in total flow. Filtration depends on how pressure is distributed between the afferent and efferent circulation, plasma flow through functioning glomeruli, and the filtering and tubular environment. Total arterial inflow does not uniquely determine those components.

Langenberg and colleagues continuously measured flow during experimentally induced hyperdynamic sepsis in sheep. Renal blood flow increased markedly while creatinine clearance fell and creatinine rose. This directly challenges the proposition that septic renal failure must always result from low total renal flow. It is evidence from a particular animal model, not proof that human septic AKI is generally a state of renal hyperperfusion. 14

Human findings vary with context. Skytte Larsson and colleagues compared eight patients with early septic shock with postoperative controls. GFR was lower, vascular resistance was higher, and total renal flow tended to be lower in the septic group. The small comparative study cannot establish one universal human pattern or identify afferent and efferent behavior from total resistance alone. 15

Several mechanisms can coexist. Regional capillary flow can become heterogeneous, inflammatory signaling can alter endothelial and tubular function, and edema or venous pressure can change the local environment. Altered tubular handling changes salt delivery to the macula densa, potentially modifying filtration through feedback. A proposed role for any one mechanism should not be mistaken for proof that it dominates every patient.

Likewise, preserved global RBF does not establish adequate oxygen delivery to every renal region or appropriate cellular use. Chapter 44 develops those oxygenation mechanisms. Here the important distinction is between measuring total inflow and explaining the function of the glomerular–tubular system.

The clinical consequence is a broader assessment of declining filtration. Hemodynamics, regional and downstream conditions, obstruction, exposures, and the trajectory of organ dysfunction remain relevant even after global cardiac output has improved.

43.11 Urine Output and Creatinine Follow Filtration Through Different Processes

Urine volume depends on filtered water minus net tubular reabsorption. At the illustrative GFR of 120 mL/min, 7,200 mL is filtered each hour. If urine output is 60 mL/h, only 60/7,200, approximately 0.83%, of that water appears as urine. Most filtered water has returned to the circulation. Small changes in fractional reabsorption can therefore produce substantial changes in urine volume without an equivalent proportional change in GFR.

Antidiuretic signaling, sodium retention, osmotic solute excretion, diuretic exposure, and obstruction all alter urine output. Low output can accompany reduced filtration or appropriate conservation of water. Preserved output can coexist with impaired filtration, especially when reabsorption is reduced. The volume remains clinically informative, but it is not a direct filtration measurement or a stand-alone distinction between functional and structural injury.

Creatinine follows another mass balance. Its concentration depends on production, distribution volume, and removal. When GFR falls abruptly, creatinine initially continues entering the body pool faster than it leaves. Serum concentration rises over time toward a new state rather than jumping immediately to a value that identifies the new GFR.

Waikar and Bonventre modeled these kinetics and demonstrated how baseline function changes the timing and percentage of the creatinine rise after a clearance reduction. Their model explains an interpretive limitation; its proposed historical classification thresholds are not presented here as current diagnostic guidance. 16

Changing production during critical illness, altered muscle mass, fluid accumulation, and tubular secretion add further uncertainty. Conventional creatinine-based eGFR assumes a sufficiently steady relationship between production and elimination. During rapid change, the reported number cannot be treated as an instantaneous measured filtration rate.

Serial creatinine and urine output should therefore be interpreted with timing, fluid balance, exposure history, and the clinical course. An early reassuring creatinine cannot exclude a recent filtration decline. An output increase after a transport-altering intervention does not independently establish recovery of filtration.

43.12 Integrating Flow, Filtration, Pressure, and Time

Consider an illustrative patient with sepsis whose cardiac output and arterial pressure improve after initial treatment. Urine output remains low, and creatinine begins rising several hours later. These observations establish an evolving renal problem, but they do not identify its vascular mechanism or prove a continuing shortage of circulating volume.

Start by separating the quantities. Improved global output does not measure renal arterial flow. Preserved total renal flow, if reliably measured, would still not establish preserved glomerular pressure or filtration. A lower GFR could reflect changes in arteriolar balance, filtering capacity, downstream pressure, or several interacting processes.

Suppose measured or appropriately estimated RPF is 400 mL/min and GFR is 80 mL/min. FF is 80/400 = 20%. Both values have fallen by one-third from the earlier illustrative values of 600 and 120, yet FF is unchanged. An apparently preserved fraction therefore cannot certify that absolute flow or filtration is satisfactory. Neither the absolute values nor the fraction establishes oxygen adequacy without further information.

Next inspect the pressure environment. A rising CVP, new abdominal distension, or an obstructed urinary catheter changes the differential. These findings are not interchangeable, and a single MAP–CVP subtraction cannot locate the responsible resistance or back pressure. Review ventilation and effort when comparing pressure measurements.

Then connect exposures and timing. Transport inhibition can change macula-densa input and urine volume. Angiotensin-dependent pressure support or prostaglandin-dependent vasodilation can alter the response to drugs. A delayed creatinine rise can reflect an earlier loss of filtration, so its timing should not automatically be assigned to the most recent intervention.

The resulting explanation should guide targeted reassessment rather than reflex escalation of pressure or volume. Follow perfusion, drainage, biochemical trajectory, and renal function while recognizing that each has a different timescale. The purpose of renal physiology is to make those differences interpretable, not to force them into one number.

References

1. Smith HW, Finkelstein N, Aliminosa L, Crawford B, Graber M. The renal clearances of substituted hippuric acid derivatives and other aromatic acids in dog and man. J Clin Invest. 1945;24(3):388–404. doi:10.1172/JCI101618

2. Brenner BM, Troy JL, Daugharty TM. The dynamics of glomerular ultrafiltration in the rat. J Clin Invest. 1971;50(8):1776–1780. doi:10.1172/JCI106667

3. Blantz RC, Konnen KS, Tucker BJ. Angiotensin II effects upon the glomerular microcirculation and ultrafiltration coefficient of the rat. J Clin Invest. 1976;57(2):419–434. doi:10.1172/JCI108293

4. Takenaka T, Suzuki H, Okada H, Hayashi K, Kanno Y, Saruta T. Mechanosensitive cation channels mediate afferent arteriolar myogenic constriction in the isolated rat kidney. J Physiol. 1998;511(Pt 1):245–253. doi:10.1111/j.1469-7793.1998.245bi.x

5. Bell PD, Lapointe JY, Sabirov R, et al. Macula densa cell signaling involves ATP release through a maxi anion channel. Proc Natl Acad Sci U S A. 2003;100(7):4322–4327. doi:10.1073/pnas.0736323100

6. Castrop H, Huang Y, Hashimoto S, et al. Impairment of tubuloglomerular feedback regulation of GFR in ecto-5′-nucleotidase/CD73-deficient mice. J Clin Invest. 2004;114(5):634–642. doi:10.1172/JCI21851

7. Sun D, Samuelson LC, Yang T, et al. Mediation of tubuloglomerular feedback by adenosine: evidence from mice lacking adenosine 1 receptors. Proc Natl Acad Sci U S A. 2001;98(17):9983–9988. doi:10.1073/pnas.171317998

8. Oppermann M, Mizel D, Huang G, et al. Macula densa control of renin secretion and preglomerular resistance in mice with selective deletion of the B isoform of the Na,K,2Cl co-transporter. J Am Soc Nephrol. 2006;17(8):2143–2152. doi:10.1681/ASN.2006040384

9. Hashimoto S, Huang Y, Briggs J, Schnermann J. Reduced autoregulatory effectiveness in adenosine 1 receptor-deficient mice. Am J Physiol Renal Physiol. 2006;290(4):F888–F891. doi:10.1152/ajprenal.00381.2005

10. Asfar P, Meziani F, Hamel JF, et al.; SEPSISPAM Investigators. High versus low blood-pressure target in patients with septic shock. N Engl J Med. 2014;370(17):1583–1593. doi:10.1056/NEJMoa1312173

11. Endo A, Yamakawa K, Tagami T, et al.; OPTPRESS Investigators. Efficacy of targeting high mean arterial pressure for older patients with septic shock (OPTPRESS): a multicentre, pragmatic, open-label, randomised controlled trial. Intensive Care Med. 2025;51(5):883–892. doi:10.1007/s00134-025-07910-4

12. Rossat J, Maillard M, Nussberger J, Brunner HR, Burnier M. Renal effects of selective cyclooxygenase-2 inhibition in normotensive salt-depleted subjects. Clin Pharmacol Ther. 1999;66(1):76–84. doi:10.1016/S0009-9236(99)70056-1

13. Harman PK, Kron IL, McLachlan HD, Freedlender AE, Nolan SP. Elevated intra-abdominal pressure and renal function. Ann Surg. 1982;196(5):594–597. doi:10.1097/00000658-198211000-00015

14. Langenberg C, Wan L, Egi M, May CN, Bellomo R. Renal blood flow in experimental septic acute renal failure. Kidney Int. 2006;69(11):1996–2002. doi:10.1038/sj.ki.5000440

15. Skytte Larsson J, Krumbholz V, Enskog A, Bragadottir G, Redfors B, Ricksten SE. Renal blood flow, glomerular filtration rate, and renal oxygenation in early clinical septic shock. Crit Care Med. 2018;46(6):e560–e566. doi:10.1097/CCM.0000000000003088

16. Waikar SS, Bonventre JV. Creatinine kinetics and the definition of acute kidney injury. J Am Soc Nephrol. 2009;20(3):672–679. doi:10.1681/ASN.2008070669