CKD Case Atlas — Companion Analysis

Critical Management Decision Nodes

Six case trajectories drawn from the source e-log corpus, isolated at the point where a clinical team had to choose a management path under real physiological constraint — and what happened after. Each card follows Trajectory → Decision → Outcome, followed by cross-case insights on the recurring management logic in these records.

Source material: de-identified student case logs (e-log format) from a general medicine / nephrology teaching archive. Presented for educational pattern review, not as validated treatment protocol. Only cases with clear, dated decision-and-response sequences in the source text are included — 6 of 156 logged cases met that bar.

The six decision nodes

Ordered by how forcing the decision was — from a diuretic ladder that needed one more rung, to a seizing, hypotensive patient where dialysis itself was the risk being weighed against the risk of not dialysing.

6
decision-node cases
4
on maintenance / new dialysis
3
hypertensive crises managed
2
transfusion-first sequencing
1
fatal-risk encephalopathy course
Hypertensive crisis on dialysis

Uncontrolled BP mid-session in a patient shifted from dialysis for respiratory distress

62F · CKD on MHD, HTN 7yr · presenting SOB grade IV, pedal edema, decreased urine output
Case #33
source ↗
Trajectory
HTN diagnosed 7 years prior on Atenolol; pedal edema and facial puffiness a year back dismissed as "due to hypertension"; underlying kidney disease only picked up on ultrasound 4 months before admission. On this admission she was shifted off dialysis mid-session to the medical unit for sweating and SOB — and during that transfer her blood pressure was recorded at 210/130 mmHg.
BP210/130SpO280%→90%ABGmild acidosis
Decision
A staged antihypertensive ladder was run in real time rather than one large bolus dose:
Nicardia 20mg + Arkamin 0.1mg given stat → BP did not fall → Lasix 80mg + Labetalol 10mg added → BP eased to 170/110 → a further Labetalol 10mg dose brought it to 160/110 → thirty minutes later 140/90 → Nitroglycerin used to consolidate control. In parallel, SpO2 was sitting at 80%, so BiPAP was started concurrently rather than waiting for the BP problem to resolve first — treating the respiratory and pressure crises as linked, not sequential.
Outcome
BiPAP brought saturation to 90% over two hours despite the ABG still showing mild acidosis; the patient was then weaned to plain oxygen and stabilised. She continued on a fixed 9-day regimen of fluid restriction (<1.5L/day), salt restriction (<2g/day), Lasix, and Nicardia — with antibiotics (Augmentin, then escalated to Levofloxacin) added from day 4 once a presumed respiratory infection component was suspected, and 3 hemodialysis sessions plus 3 blood transfusions run across the admission.
Why this node mattersA hypertensive emergency arriving mid-dialysis-session forces a choice between finishing ultrafiltration (which itself drops pressure unpredictably) and switching to pharmacologic titration. The team titrated in small graded steps (four drug adjustments in roughly an hour) rather than one aggressive dose — the safer approach when a dialysis patient's volume status is already in flux.
Dialysis-vs-hypotension trade-off

Seizure and coma from uremic encephalopathy — dialysing a patient already in shock

48M · undiagnosed CKD stage V, denovo HTN + DM, NSAID abuse 2yr · new-onset tonic seizure, then comatose
Case #646
source ↗
Trajectory
A farmer, CKD flagged incidentally at a femur-fracture operation two years earlier but lost to follow-up, presented after a witnessed bilateral-limb tonic seizure with tongue bite and ten hours of post-ictal confusion. On admission: creatinine 9.3, urea 161, denovo diabetes, and a working diagnosis of severe metabolic acidosis with uremic encephalopathy. By day 1 he had deteriorated to comatose (GCS E1V1M1, fixed mid-dilated pupils, absent deep tendon reflexes).
BP100/40 (on pressors)PR107SpO294% on 8L O2Creat9.3
Decision
Dialysis was still performed despite a MAP low enough to require triple inotrope/vasopressor support:
Noradrenaline, dobutamine and vasopressin infusions were started to hold blood pressure, and — rather than deferring dialysis until hemodynamically stable — the team ran one session of hemodialysis with a deliberately low ultrafiltration rate "due to his hypotension," alongside 2 units of PRBC and 300mEq of bicarbonate replaced intra-dialysis. The refractory acidosis itself was treated as the driver of the shock, so clearing it was judged to outweigh the risk of further hypotension from fluid removal — hence UF was throttled rather than dialysis withheld.
Outcome
The seizures were recorded as resolved, but the acidosis was explicitly documented as "refractory" and the shock state persisted, requiring continued vasopressor titration to MAP targets. The team's own open question in the notes — "can acidosis cause shock without decreasing myocardial contractility?" — signals that the outcome at last documented note remained unresolved/guarded rather than a clean recovery.
Why this node mattersThis is the corpus's clearest example of a genuine equipoise decision: standard teaching would delay dialysis in hemodynamic shock, but refractory uremic acidosis was judged the proximate threat. The mitigation wasn't "dialyse or don't" — it was a third option, low-UF dialysis under pressor cover, that tried to get the metabolic benefit while limiting the hemodynamic cost.
Sequencing transfusion before BP control

Hypertensive emergency in a patient with a hemoglobin of 4 g/dl

40F · CKD + dilated cardiomyopathy, denovo hypothyroidism · referred in with Hb 4g%, platelets 25k
Case #739
source ↗
Trajectory
A young woman with unclear glomerulonephritis (near-normal renal biopsy) and known DCMP presented after a local hospital found Hb 4g/dl and platelets 25k, treated with erythropoietin and 5 units PRBC before transfer. On arrival at this facility she carried a provisional diagnosis of hypertensive emergency with grade 1 retinopathy, on a background of raised JVP and a parasternal heave.
BP160/100PR125SpO288%Hb (post-transfusion)7.3
Decision
Correction of the anemia was front-loaded at the referring hospital before antihypertensive titration began here:
Transfusing to raise Hb from 4 to roughly 7.3 g/dl before intensifying antihypertensive therapy reduces the risk that vasodilating/negative-inotrope agents would be started on top of severe anemia-driven high-output physiology (tachycardia 125bpm, raised JVP) — a combination that can precipitate cardiovascular collapse. Once she arrived here already transfused, the regimen started conservatively: Lasix, Telmisartan, fluid/salt restriction, with 4th-hourly vitals — not an aggressive multi-agent push given her still-fragile cardiac state.
Outcome
By the next day, BP had come down from 160/100 to 120/90 on the conservative regimen alone (Lasix + Met-XL + Nicardia added), and she was escalated modestly with Metolazone-class add-ons documented for cardiac support (Ecosprin, Spironolactone, Nicardia retard) rather than aggressive vasodilation — consistent with treating this as a volume-and-anemia-driven pressure problem rather than a pure vascular hypertensive emergency.
Why this node mattersThe sequencing — correct the oxygen-carrying capacity first, then titrate BP medication — is a recurring but implicit theme across the severe-anemia cases in this corpus: aggressive antihypertensive therapy on top of profound anemia risks worsening high-output cardiac strain rather than relieving it.
Diuretic-resistant anasarca

Escalating from bolus Lasix to a continuous infusion, then adding metolazone sequential blockade

45M · AKI on CKD (diabetic nephropathy), COPD, anemia of chronic disease · gross ascites, scrotal edema, fluid thrill
Case #717
source ↗
Trajectory
Progressive pedal edema over a month advanced to ascites with a positive fluid thrill and new scrotal swelling. On admission: BP 180/100, creatinine 3.8, urea 85, INR 1.9 — decompensated fluid overload not responding to home management.
Day 1Lasix 80mg IV statDay 2Lasix drip 20mg/hrDay 5HD + 1 PRBC
Decision
The diuretic strategy escalated in three distinct steps rather than repeating the same bolus dose:
Step 1 (day 1): Lasix 80mg IV stat plus Metolazone 5mg to pre-block distal sodium reabsorption. Step 2 (day 2): converted from intermittent boluses to a continuous Lasix infusion at 20mg/hr — the standard response to diminishing returns from bolus dosing in resistant fluid overload. Step 3 (day 3–5): with ascites still not clearing and hypoxia (end-inspiratory wheeze bilaterally) persisting, the team added hemodialysis with a unit of PRBC for concurrent anemia — using ultrafiltration as a direct mechanical solution once pharmacologic diuresis plateaued.
Outcome
BP responded (180/100 → 120/60 by day 3) but then rebounded to 170/70–170/90 by days 6–7, and a fever spike appeared alongside worsening bilateral pedal edema — prompting a further escalation to empiric IV vancomycin through a central line, suggesting a suspected catheter-related or systemic infection superimposed on the fluid-overload course rather than resolution.
Why this node mattersThis case documents the textbook diuretic-resistance ladder end to end inside one record: thiazide-type sequential blockade → bolus-to-infusion conversion → ultrafiltration — and shows what triggers the next rung (persistent effusion/wheeze, not just the edema itself).
Infection + incidental GI bleed

Antibiotic choice for CAP-on-CKD, complicated by a melena episode mid-course

80M · CKD 2yr on conservative management, past pulmonary TB, HTN 2yr · fever + SOB + loose stools, SpO2 90% RA
Case #1060
source ↗
Trajectory
A patient who had self-discontinued renal medication a year earlier (staying only on Nodosis) presented in type 1 respiratory failure with fever/chills, an elevated TLC of 22,400, and a chest picture read as community-acquired pneumonia layered onto known CKD — with a normal creatinine of 1.2 at this admission, distinguishing it from a purely renal crisis.
SpO290% RATLC22,400pO2 (ABG)57.2Creat1.2
Decision
Empiric dual antibiotic cover was chosen up front, and fluid targets were kept renal-conservative despite the primary problem being respiratory:
IV Augmentin + oral Azithromycin were started together on admission (rather than a single agent) — reasonable for a CKD patient where narrower cultures take time and the TLC/fever picture suggested a more severe CAP. Fluid restriction (<1L/day) and salt restriction (<2.4g/day) were kept in place throughout the pneumonia treatment, showing the CKD baseline was not relaxed just because the acute problem was respiratory. When black tarry stools appeared on day 2 — after a recent Dulcolax suppository — the team's response was monitoring rather than immediately stopping antibiotics or adding empiric PPI escalation, judged most likely a local mucosal effect rather than an antibiotic-driven bleed.
Outcome
SOB and cough progressively decreased over days 2–4, SpO2 rose from 90% to 99–100% on titrated oxygen (weaned from BiPAP-level support to 8L then room air over 4 days), and the melena did not recur — supporting the working call that it was self-limited rather than a marker to change the antibiotic regimen. The same core regimen (Augmentin, Azithromycin, Amlodipine, Nodosis) ran unchanged for all 4 documented days, with only oxygen delivery and CPAP frequency titrated down as the response confirmed itself.
Why this node mattersThe decision not to escalate or change antibiotics after an ambiguous bleeding signal — while still watching closely — reflects a judgment that changing a working regimen on an unconfirmed signal carries its own risk, particularly in a CKD patient where drug substitution options are already narrowed by renal clearance.
Fluid target vs. renal reserve

Loosening the fluid-restriction ceiling as a marker of clinical improvement

60M · CKD, on Nicardia/Nodosis/Erythropoietin maintenance · abdominal distension, no fever spikes on review
Case (CKD ward, 4/1/2022)
source ↗
Trajectory
A CKD ward patient stable on maintenance therapy (Nicardia, Nodosis, Shelcal, Orofer, weekly erythropoietin) with persistent non-tender abdominal distension but no fever, stable BP (130/80) and pulse (88bpm) at soap-note review.
BP130/80PR88Fluid limit (prior)<1L/dayFluid limit (revised)<2L/day
Decision
The fluid restriction ceiling was relaxed from <1L/day to <2L/day at the same visit medications otherwise stayed unchanged:
With vitals stable, no fresh complaints, and no fever spikes, the team used the fluid allowance itself as an adjustable dial rather than a fixed rule — loosening it once the abdominal distension was judged non-tender and non-progressive, while keeping every drug (Nicardia, Nodosis, Shelcal, Orofer, erythropoietin) at the same dose. Ultracet and nebulization (Budecort/Duolin) were continued alongside, suggesting a comorbid respiratory component was still being actively managed even as renal fluid targets eased.
Outcome
The record shows the patient tolerating the wider fluid allowance without new complaints in the documented follow-up, with vitals and abdominal exam unchanged — a pragmatic signal that the tighter restriction had been a precaution during instability rather than a permanent floor.
Why this node mattersMost cases in this corpus show restriction tightening as decompensation is treated; this is the rarer example of the reverse move — restriction eased as a direct response to sustained stability, illustrating that fluid targets in these records function as a live titration parameter, not a static prescription.

Cross-case management insights

Patterns in how decisions were made and re-made across these six trajectories, read across the source material rather than from any single case.

01

Crises are titrated in small steps, not single large doses

In every hypertensive-crisis node (#33, #717, #739), BP was brought down through a sequence of small drug additions with a recheck between each — never one aggressive bolus. In a population where volume status is already unstable from renal failure, this graded approach avoids overshooting into hypotension on top of dialysis-driven fluid shifts.

Evidence: #33 (4-step titration in ~1hr), #717 (Lasix→Metolazone→infusion), #739 (transfusion before BP drugs)
02

Anemia correction is sequenced ahead of vasoactive therapy

Where severe anemia (Hb 4–7g/dl) coincided with hypertension or cardiac strain, transfusion consistently preceded or ran alongside — rather than after — antihypertensive escalation, likely to avoid compounding high-output cardiac stress with vasodilation.

Evidence: #739 (5 PRBC before BP titration), #646 (PRBC given during the same dialysis session as pressor support)
03

Dialysis timing is weighed against hemodynamic tolerance, not deferred by rule

In the most severe node (#646), dialysis was performed despite shock-range blood pressure by pairing it with vasopressor support and a deliberately reduced ultrafiltration rate — treating "when metabolic derangement itself is driving instability" as a case for dialysing carefully, not for waiting.

Evidence: #646 (low-UF HD under triple-pressor cover)
04

Diuretic resistance follows a visible three-rung ladder

Bolus loop diuretic → distal sodium blockade (metolazone) added → conversion to continuous infusion → mechanical ultrafiltration via dialysis. The trigger to climb each rung was a persistent physical sign (ascites, wheeze, fluid thrill) rather than a fixed time interval.

Evidence: #717 (full ladder across 5 documented days)
05

Ambiguous new signals mid-course are monitored, not reflexively treated

A melena episode appearing during antibiotic therapy for CAP-on-CKD did not trigger a regimen change; the team continued the working antibiotic combination and watched for recurrence, weighing the disruption of an effective plan against an unconfirmed adverse signal.

Evidence: #1060 (Dulcolax-associated melena, day 2)
06

Fluid restriction functions as a live dial, not a fixed rule

Restrictions were tightened during instability and loosened once vitals and exam stabilised — the only case of "de-escalation" documented among these six, and a reminder that conservative CKD targets are not meant to be permanent ceilings once the acute driver resolves.

Evidence: CKD ward case, <1L/day → <2L/day after 2 stable review days

Composite diuretic-resistance ladder (synthesized from #717 & #33)

Bolus loop diuretic (Lasix 40–80mg IV)→ no response →
Add distal blockade (Metolazone 5mg)→ no response →
Convert to continuous infusion (Lasix 20mg/hr)→ effusion / thrill persists →
Mechanical ultrafiltration via hemodialysis→ reassess for infective driver if fever/rebound BP →
Broaden antimicrobial cover (e.g. empiric vancomycin)