Clinical spectrum and hospital course of acute kidney injury in elderly patients at a tertiary care hospital

Title

Clinical Spectrum and Hospital Course of Acute Kidney Injury in Elderly Patients at a Tertiary Care Hospital: A Hospital-Based Prospective Observational Study

• Speciality: MD General Medicine
• Subject area: Nephrology and Electrolytes
• Study design: Hospital-based prospective observational (descriptive with analytical component) study

Introduction

Acute kidney injury (AKI) is an abrupt decline in kidney function, measured by a rise in serum creatinine, a fall in urine output, or both, over hours to days. The Kidney Disease: Improving Global Outcomes (KDIGO) 2012 guideline combined the earlier RIFLE and AKIN systems into a single definition and a three-stage severity classification. This is now the standard used in clinical practice and research.

Elderly patients are especially prone to AKI. With age, the kidney undergoes structural and functional change:

• Fewer functioning nephrons, glomerulosclerosis and interstitial fibrosis, so renal functional reserve is reduced.
• Weaker renal autoregulation, so the kidney tolerates less hypotension and volume depletion.
• Poorer urinary concentrating ability and impaired thirst, so elderly patients become dehydrated easily.
• Tubular dysfunction and slower repair after injury, so recovery is slower and less complete.

Several other factors add to this risk: comorbidities such as diabetes, hypertension, heart failure and chronic kidney disease (CKD); polypharmacy, including NSAIDs, ACE inhibitors or ARBs, diuretics and nephrotoxic antibiotics; frequent exposure to iodinated contrast; obstructive uropathy from prostatic enlargement or malignancy; and a high burden of sepsis.

AKI in the elderly is also harder to recognise. Because of sarcopenia, an older patient with low muscle mass can have a "normal-looking" serum creatinine despite markedly reduced GFR. This delays diagnosis and leads to under-staging. In India, the picture is further shaped by community-acquired AKI from diarrhoeal illness, tropical infections (malaria, leptospirosis, scrub typhus, dengue), over-the-counter analgesic use and late presentation. These sit alongside the hospital-acquired AKI seen in ageing populations worldwide.

Describing the clinical spectrum of AKI in elderly patients admitted to an Indian tertiary care hospital (its causes, setting, severity and in-hospital course) can guide early recognition, nephrotoxin stewardship, triage for kidney replacement therapy and counselling of families.

Need for the Study

• Most Indian AKI series include all adults. Elderly patients are often a subgroup, so age-specific aetiology, severity and outcome are not described in enough detail for local practice.
• The share of AKI caused by sepsis, volume depletion, nephrotoxic drugs, obstruction and tropical infection differs by region and by referral pattern. Institution-specific data are needed to plan preventive protocols.
• Elderly patients have higher rates of AKI-on-CKD, multifactorial AKI and incomplete recovery. These need to be documented with standardised KDIGO definitions and an explicit rule for baseline creatinine.
• Frailty, comorbidity burden and polypharmacy are rarely recorded systematically in Indian AKI studies, even though they are routine in geriatric medicine.
• Prospective data on recovery status at discharge, need for dialysis and in-hospital mortality will help with prognostic counselling and resource planning (dialysis slots, ICU beds) in the department.
• The study needs no extra intervention and uses investigations already done as standard care, so it is feasible within an MD residency.

Review of Literature

Definitions and evolution of AKI criteria

• AKI definitions developed from varied creatinine thresholds to the RIFLE criteria, then AKIN, and finally the unified KDIGO 2012 criteria. KDIGO is now accepted as the reference standard for diagnosis and staging.
• The literature highlights the problem of baseline creatinine when no pre-admission value is available. Different methods are used: the lowest in-hospital value, back-calculation from an assumed eGFR, or the admission value. Each can misclassify AKI, and back-calculation is especially unreliable in elderly patients and in those with unrecognised CKD.
• The urine output criterion is often under-used outside ICUs because hourly measurement is not available. Studies that use only creatinine may under-estimate AKI incidence.

The ageing kidney and susceptibility to AKI

• Age-related decline in GFR, loss of renal reserve, reduced autoregulation and slower tubular regeneration are described as the biological basis for higher AKI incidence and poorer recovery in older people.
• Serum creatinine depends on muscle mass. Reviews caution that creatinine-based equations can over-estimate kidney function in frail, sarcopenic elderly patients.

Aetiology and clinical spectrum

• Studies from high-income countries describe hospital-acquired AKI dominated by sepsis, major surgery, cardiorenal syndrome, contrast exposure and drug toxicity.
• Studies from India and other tropical, low- and middle-income settings describe a higher share of community-acquired AKI, linked to volume depletion from gastroenteritis, infections including tropical fevers, envenomation, and drug or herbal toxicity. Their patients are generally younger than in Western cohorts.
• In elderly patients, the literature consistently describes AKI as multifactorial: prerenal causes, sepsis-associated acute tubular necrosis, nephrotoxic medicines (including the "triple whammy" of NSAID, renin-angiotensin blocker and diuretic) and obstructive uropathy all contribute.

Hospital course and outcomes

• AKI is consistently associated with higher in-hospital mortality, longer hospital stay, need for kidney replacement therapy and, among survivors, progression to CKD. The risk rises with higher KDIGO stage.
• Factors repeatedly reported as associated with poor outcome include sepsis, oliguria, need for mechanical ventilation or vasopressors, higher severity scores, multi-organ dysfunction and pre-existing CKD.
• Decisions about dialysis in very elderly and frail patients are discussed in the literature, together with the role of frailty scores and goals-of-care conversations.
• Acute kidney disease (AKD), meaning kidney dysfunction lasting 7–90 days after AKI, and incomplete recovery at discharge are recognised as markers of long-term risk.

Gap identified

• There are few prospective Indian studies limited to patients aged 60 years and above that use KDIGO staging, a stated baseline-creatinine rule, and frailty and comorbidity scores, and that report recovery status at discharge.

Suggested PubMed search strings

The resident should run these searches, read the actual papers, and cite only studies they have read:

1. ("acute kidney injury"[MeSH Terms] OR "acute kidney injury"[tiab]) AND ("aged"[MeSH Terms] OR elderly[tiab]) AND (etiology[tiab] OR spectrum[tiab] OR profile[tiab])
2. "acute kidney injury"[tiab] AND (elderly[tiab] OR "older adults"[tiab] OR geriatric[tiab]) AND India[tiab]
3. "acute kidney injury"[tiab] AND KDIGO[tiab] AND "aged"[MeSH Terms] AND (mortality[tiab] OR outcome[tiab])
4. "community-acquired"[tiab] AND "acute kidney injury"[tiab] AND (India[tiab] OR tropical[tiab] OR "developing countries"[MeSH Terms])
5. "acute kidney injury"[tiab] AND ("renal recovery"[tiab] OR "kidney recovery"[tiab]) AND "aged"[MeSH Terms]
6. "acute kidney injury"[tiab] AND (frailty[MeSH Terms] OR frailty[tiab]) AND hospitali*[tiab]
7. "baseline creatinine"[tiab] AND "acute kidney injury"[tiab] AND (definition[tiab] OR estimation[tiab] OR misclassification[tiab])
8. "hospital-acquired"[tiab] AND "acute kidney injury"[tiab] AND (elderly[tiab] OR "aged"[MeSH Terms]) AND "risk factors"[MeSH Terms]

Research Question and Hypothesis

Research question

In patients aged 60 years and above admitted to the Department of General Medicine of a tertiary care hospital with AKI, or developing AKI during admission:

• What are the aetiology, setting of onset and KDIGO severity of AKI?
• What is their hospital course in terms of renal recovery, need for kidney replacement therapy and in-hospital mortality?
• Which clinical factors are associated with in-hospital death?

Hypothesis

The primary objective is descriptive and has no formal hypothesis. The hypotheses below apply to the main analytical secondary objective.

• Null hypothesis (H0): In elderly patients with AKI, in-hospital mortality does not differ by maximum KDIGO stage. Sepsis, frailty (Clinical Frailty Scale ≥5) and pre-existing CKD are not independently associated with in-hospital mortality.
• Alternative hypothesis (H1): In elderly patients with AKI, in-hospital mortality increases with maximum KDIGO stage. At least one of sepsis, frailty or pre-existing CKD is independently associated with in-hospital mortality.

Aims and Objectives

Aim

To study the clinical spectrum and hospital course of acute kidney injury in patients aged 60 years and above admitted to a tertiary care hospital.

Primary objective

To describe, in patients aged ≥60 years with KDIGO-defined AKI:

• the aetiological profile of AKI;
• whether AKI was community- or hospital-acquired;
• the maximum KDIGO stage; and
• the in-hospital outcome: renal recovery status at discharge (complete recovery, partial recovery, dialysis dependence) and in-hospital mortality.

Secondary objectives

1. To compare the aetiology, severity and outcome of AKI across age groups (60–69, 70–79 and ≥80 years).
2. To compare the clinical profile and outcome of community-acquired versus hospital-acquired AKI.
3. To identify clinical and laboratory factors associated with in-hospital mortality, including maximum KDIGO stage, sepsis, oliguria, SOFA score, Clinical Frailty Scale, Charlson Comorbidity Index and pre-existing CKD.
4. To determine the proportion of patients needing kidney replacement therapy (KRT), the indications for it, and the length of hospital stay by KDIGO stage.
5. To determine the proportion of patients with nephrotoxic medication exposure and polypharmacy before the onset of AKI.

Materials and Methods

Study design

Hospital-based prospective observational study (descriptive, with an analytical component).

Study setting

The Department of General Medicine (medical wards and Medical Intensive Care Unit) of [Name of Institution], a tertiary care teaching hospital in [City, State]. Patients managed jointly with the Department of Nephrology (dialysis unit) are included while they stay under General Medicine.

Study duration

24 months from the date of Institutional Ethics Committee (IEC) approval:

• 18 months of patient recruitment and follow-up;
• 6 months for preparation, data analysis and thesis writing.

Study population

Patients aged 60 years and above admitted to the Department of General Medicine who meet the KDIGO 2012 criteria for AKI at admission or at any time during the hospital stay.

Inclusion criteria

1. Age ≥60 years (completed years, as per hospital records or a valid identity document).
2. Admitted to General Medicine wards or the Medical ICU.
3. Meeting KDIGO 2012 criteria for AKI (see operational definitions), either at admission or during hospital stay.
4. Written informed consent from the patient or, if the patient lacks capacity, from the legally acceptable representative (LAR).

Exclusion criteria

1. Known CKD stage G4–G5 (eGFR <30 mL/min/1.73 m² for >3 months) or patients on maintenance dialysis.
2. Kidney transplant recipients.
3. Patients with fewer than two serum creatinine values during the stay, which makes AKI impossible to confirm or stage (for example, death or discharge within 24 hours of admission).
4. Patients transferred out to another hospital within 48 hours of enrolment, so the hospital course cannot be followed.
5. Refusal of consent, or no LAR available for a patient who lacks capacity.

Patients with known CKD stage G1–G3 are included and analysed as an AKI-on-CKD subgroup.

Sample size

Basis: The primary objective is descriptive. The sample size is therefore calculated to estimate a single proportion with acceptable absolute precision. In-hospital mortality is used as the anchor proportion because it is the most clinically important outcome in the hospital course.

Formula (single proportion):

n = Z² × p × (1 − p) / d²

Inputs:

• Z = 1.96, for a 95% confidence level.
• p = 0.25 (25%), the expected in-hospital mortality among elderly patients with AKI. ASSUMPTION: the resident must replace this with the figure reported in a published study of elderly AKI that they have read and cited, or with the mortality seen in the first 30 patients of a pilot, or with the department's previous-year records.
• d = 0.07 (7%), the absolute precision. ASSUMPTION: a choice of acceptable precision. It should be agreed with the guide and the statistician.
• Anticipated attrition = 10%, covering discharge against medical advice, transfer out and incomplete records. ASSUMPTION.

Arithmetic:

• Z² = 1.96 × 1.96 = 3.8416
• p × (1 − p) = 0.25 × 0.75 = 0.1875
• Z² × p × (1 − p) = 3.8416 × 0.1875 = 0.7203
• d² = 0.07 × 0.07 = 0.0049
• n = 0.7203 / 0.0049 = 147.0, so 147 patients
• Adjusted for 10% attrition: 147 / (1 − 0.10) = 147 / 0.90 = 163.3, rounded up to 164 patients

Final sample size: 164 elderly patients with AKI.

Sensitivity of the calculation (redo the calculation once the real figure is known):

• If the cited or pilot mortality is 30%: n = 3.8416 × 0.30 × 0.70 / 0.0049 = 0.8067 / 0.0049 = 164.6, giving 165; with 10% attrition, 165 / 0.90 = 183.3, giving 184.
• If the committee wants every proportion (including aetiology categories) estimated to ±7%, use the conservative p = 0.50: n = 3.8416 × 0.25 / 0.0049 = 196.0, giving 196; with 10% attrition, 196 / 0.90 = 217.8, giving 218.

Feasibility check: 164 patients over 18 months means about 9–10 enrolments per month. Before submitting, the resident should count the elderly AKI admissions in the department over the previous 12 months, using laboratory creatinine records or the medical records department, to confirm this rate is achievable.

Note on the analytical objective: With about 40 expected deaths (25% of 164), multivariable logistic regression can reliably include only about 4 predictors, using the rule of roughly 10 outcome events per variable. The predictors are therefore fixed in advance (see the Statistical Analysis Plan).

Sampling technique

Consecutive (non-probability) sampling. Every eligible patient is enrolled in order of identification until the required sample size is reached.

Methodology and data collection

1. Screening

• Every day, the investigator screens all General Medicine admissions aged ≥60 years and reviews their serum creatinine reports in the hospital laboratory system.
• Community-acquired AKI: patients who meet KDIGO criteria at admission or within 48 hours of admission.
• Hospital-acquired AKI: patients without AKI at admission whose creatinine rises to KDIGO criteria later in the stay. They are identified by reviewing serial creatinine values.

2. Enrolment

• Eligibility is confirmed against the inclusion and exclusion criteria.
• Written informed consent is taken from the patient or the LAR.
• A unique study ID is assigned.

3. Baseline assessment (within 24 hours of enrolment)

• Demographics.
• Presenting complaints and duration of symptoms before admission.
• Comorbidities and the Charlson Comorbidity Index.
• Clinical Frailty Scale, based on the patient's status 2 weeks before the current illness, as reported by the patient or family.
• Full drug history over the preceding 7 days, including over-the-counter, herbal and indigenous medicines.
• Recent exposure to iodinated contrast.
• Vital signs, volume status and examination findings.
• SOFA score. APACHE II is also recorded for ICU patients.

4. Aetiological work-up

• Done as per standard care (see Investigations).
• The final aetiology is assigned by the investigator and confirmed by the guide at discharge, using the operational definitions.
• When more than one cause contributes, all are recorded and the predominant cause is identified.

5. Daily follow-up until discharge, death or discharge against medical advice (LAMA)

• Serum creatinine and electrolytes are recorded as per standard care. At minimum, values from days 1, 2, 3, 5, 7, weekly thereafter, and at discharge are captured.
• Urine output is recorded wherever accurately charted.
• Fluid balance.
• KDIGO stage on each day, and the maximum stage reached.
• Complications: hyperkalaemia, metabolic acidosis, fluid overload or pulmonary oedema, uraemic encephalopathy, and hospital-acquired infection.
• ICU transfer, mechanical ventilation and vasopressor use.
• KRT: whether started, indication, modality and number of sessions.
• Nephrotoxic drugs stopped or dose-adjusted.

6. Outcome at discharge

• Recovery status (complete, partial, or dialysis-dependent), serum creatinine at discharge, length of hospital stay, and in-hospital death with its cause.
• LAMA is recorded as a separate category.

7. Treatment

• All treatment decisions rest with the treating unit. The study does not change management.

8. Data recording

• Data are entered on a pre-designed case record form (CRF).
• They are transferred within 48 hours to a password-protected spreadsheet using study IDs only.
• The guide checks 10% of CRFs against source documents each month.

Study variables and operational definitions

Elderly

• Age ≥60 years, in line with the definition of older persons used in Indian national policy.
• Age groups: 60–69, 70–79 and ≥80 years.

Acute kidney injury (KDIGO 2012)

Any one of the following:

• An increase in serum creatinine of ≥0.3 mg/dL within 48 hours.
• An increase in serum creatinine to ≥1.5 times baseline, known or presumed to have occurred within the prior 7 days.
• Urine output <0.5 mL/kg/h for 6 hours.

KDIGO staging (the higher of the creatinine or urine output criterion is used)

• Stage 1: creatinine 1.5–1.9 times baseline, or a rise of ≥0.3 mg/dL; or urine output <0.5 mL/kg/h for 6–12 hours.
• Stage 2: creatinine 2.0–2.9 times baseline; or urine output <0.5 mL/kg/h for ≥12 hours.
• Stage 3: creatinine ≥3.0 times baseline, or an increase to ≥4.0 mg/dL, or start of KRT; or urine output <0.3 mL/kg/h for ≥24 hours, or anuria for ≥12 hours.
• Maximum KDIGO stage: the highest stage reached at any time during the hospital stay.

Urine output criterion

• Applied only where urine output is accurately measured, that is, in catheterised patients or patients with charted hourly or 6-hourly output.
• In all other patients, diagnosis and staging use creatinine alone. This is stated as a limitation.

Baseline serum creatinine (fixed order of preference)

1. The most recent stable outpatient value from 7–365 days before admission.
2. If none is available, the lowest value during the current hospital stay.

Back-calculation from an assumed eGFR is not used, because it is unreliable in elderly patients. The method used for each patient is recorded.

Community-acquired AKI

• AKI present at admission, or meeting criteria within 48 hours of admission.

Hospital-acquired AKI

• AKI first meeting criteria more than 48 hours after admission, in a patient without AKI at admission.

Pre-existing CKD and AKI-on-CKD

• CKD: a documented eGFR <60 mL/min/1.73 m² (CKD-EPI 2021 equation) for >3 months, or documented structural kidney disease, before the current admission.
• AKI-on-CKD: AKI occurring in a patient with pre-existing CKD stage G1–G3.

Oliguria

• Urine output <400 mL/day, or <0.5 mL/kg/h for ≥6 hours.

Aetiological categories

• Prerenal (volume-responsive): volume depletion from vomiting, diarrhoea, poor intake or diuretics, or haemodynamic causes, with recovery of creatinine within 72 hours of volume repletion or haemodynamic correction. Supported by fractional excretion of sodium (FENa) <1%, or fractional excretion of urea (FEUrea) <35% in patients on diuretics.
• Sepsis-associated AKI: AKI in a patient with sepsis by Sepsis-3 criteria, that is, suspected or confirmed infection with an acute rise in SOFA score of ≥2 points.
• Septic shock: sepsis needing vasopressors to keep mean arterial pressure ≥65 mmHg, with serum lactate >2 mmol/L despite adequate fluid resuscitation.
• Nephrotoxic (drug-induced) AKI: AKI within 7 days of starting or increasing a nephrotoxic drug (for example NSAIDs, aminoglycosides, vancomycin, amphotericin B, or ACE inhibitor/ARB with a diuretic), with no better explanation.
• Contrast-associated AKI: AKI within 48–72 hours of intravascular iodinated contrast, with no other clear cause.
• Cardiorenal AKI: AKI with acute decompensated heart failure or acute coronary syndrome as the main precipitant.
• Hepatorenal-type AKI (HRS-AKI): AKI in cirrhosis with ascites meeting International Club of Ascites criteria.
• Obstructive (postrenal) AKI: hydronephrosis or a significant post-void residual urine on ultrasound, with improvement in creatinine after relief of the obstruction.
• Rhabdomyolysis-associated AKI: creatine kinase >5 times the upper limit of normal, with a compatible clinical setting.
• Tropical infection-associated AKI: AKI with laboratory-confirmed malaria, leptospirosis, scrub typhus or dengue.
• Glomerular, interstitial or vascular disease: suggested by active urinary sediment (dysmorphic red cells, red cell casts), significant proteinuria, systemic disease or serology, and confirmed by biopsy where the treating team performs one.
• Multifactorial: two or more of the above contributing. The predominant cause is recorded.

Comorbidity, frailty and medication burden

• Polypharmacy: regular use of ≥5 medicines before admission.
• Frailty: Clinical Frailty Scale score ≥5, based on status 2 weeks before admission.
• Comorbidity burden: Charlson Comorbidity Index (age-adjusted), as a continuous score, and categorised as 0–2, 3–4 and ≥5.

Kidney replacement therapy (KRT)

• Haemodialysis, sustained low-efficiency dialysis or peritoneal dialysis started for AKI.
• The indication is recorded: refractory hyperkalaemia, refractory metabolic acidosis, fluid overload unresponsive to diuretics, uraemic complications, or progressive azotaemia.

Renal recovery status at discharge

• Complete recovery: serum creatinine back to <1.5 times baseline (no longer meeting AKI criteria) and not dialysis-dependent.
• Partial recovery: serum creatinine still ≥1.5 times baseline but not dialysis-dependent.
• Non-recovery: dialysis-dependent at discharge.

Other outcome definitions

• AKI progression: an increase of at least one KDIGO stage after enrolment.
• Length of hospital stay: days from admission to discharge or death. Days from AKI diagnosis to discharge are recorded separately.
• In-hospital mortality: death from any cause during the index admission.
• LAMA: discharge against medical advice. Recorded separately and handled as described in the Statistical Analysis Plan.

Investigations

All investigations are part of standard care for AKI in the department. No investigation is done only for research.

At enrolment

• Complete blood count and peripheral smear.
• Serum creatinine, blood urea, sodium, potassium, chloride, calcium, phosphate and uric acid.
• Arterial or venous blood gas with bicarbonate, and serum lactate where sepsis is suspected.
• Random blood glucose and HbA1c.
• Liver function tests.
• Urine routine and microscopy, including sediment for casts and dysmorphic red cells.
• Urine spot protein-to-creatinine ratio.
• Urine sodium, urine creatinine and urine urea, to calculate FENa and FEUrea (sent before diuretics wherever possible).
• Ultrasound of the kidneys, ureters and bladder: kidney size, echotexture and corticomedullary differentiation, hydronephrosis, and post-void residual volume.
• 12-lead ECG for features of hyperkalaemia.

When clinically indicated

• Blood and urine cultures.
• Procalcitonin.
• Malaria antigen and smear, leptospira IgM, scrub typhus IgM, and dengue NS1/IgM.
• Creatine kinase.
• 2D echocardiography.
• Chest X-ray.
• Serum protein electrophoresis and free light chains (if myeloma is suspected).
• ANA, ANCA, complement levels and anti-GBM (if glomerulonephritis is suspected).
• Kidney biopsy, only if the treating team decides it is needed.

Serial

• Serum creatinine and electrolytes as per the follow-up schedule above, and at discharge.

Statistical Analysis Plan

Software and data handling

• Data are entered in Microsoft Excel and analysed in IBM SPSS Statistics (or the free alternatives R or Jamovi). The version used is stated in the thesis.
• Before analysis, data are cleaned: range checks, logic checks (for example, maximum stage ≥ enrolment stage) and resolution of missing values against source documents.

Descriptive statistics

• Categorical variables (aetiology, community- or hospital-acquired onset, KDIGO stage, recovery status, mortality) are shown as frequencies and percentages with 95% confidence intervals (Wilson method).
• Continuous variables are tested for normality with the Shapiro–Wilk test and Q-Q plots.
• Normally distributed data are shown as mean ± SD; skewed data (creatinine, length of stay) as median (IQR).

Comparisons (secondary objectives 1, 2 and 4)

• Categorical variables across groups: Chi-square test, or Fisher's exact test when any expected cell count is <5.
• Trends across ordered groups (age groups, KDIGO stages): Chi-square test for trend.
• Continuous variables between two groups (community- vs hospital-acquired; survivors vs non-survivors): independent t-test or Mann–Whitney U test.
• Continuous variables across three groups (age groups, KDIGO stages): one-way ANOVA with post-hoc Tukey test, or Kruskal–Wallis test with Dunn's post-hoc test and Bonferroni correction.

Factors associated with in-hospital mortality (secondary objective 3)

• Univariable analysis: each candidate factor is compared between survivors and non-survivors. Unadjusted odds ratios with 95% CIs are reported.
• Multivariable binary logistic regression, using pre-specified predictors only to respect the events-per-variable limit:
• maximum KDIGO stage 3 (vs stages 1–2);
• sepsis;
• Clinical Frailty Scale ≥5;
• pre-existing CKD.
• SOFA score and the Charlson Index are not entered alongside these predictors, to avoid overlap. They are examined in a separate model if there are enough deaths.
• Results are reported as adjusted odds ratios with 95% CIs.
• Model checks: multicollinearity (variance inflation factor), the Hosmer–Lemeshow goodness-of-fit test, and Nagelkerke R².
• Optional: the ROC curve and area under the curve for SOFA score in predicting mortality.

Handling LAMA and missing data

• Primary analysis: patients discharged against medical advice are excluded from the recovery and mortality denominators. Their numbers and baseline characteristics are reported.
• Sensitivity analysis: patients discharged against medical advice are counted as deaths (worst case), and the results are compared.
• Variables with more than 10% missing data are reported as such. No imputation is done.

Significance level

• A two-tailed p < 0.05 is considered statistically significant.

Presentation

• Results are presented as text and charts: bar charts for aetiology and stage, a stacked bar chart of outcome by KDIGO stage, and a box plot of length of stay.
• A STROBE-style flow diagram shows patients screened, eligible, enrolled and analysed.
• Reporting follows the STROBE checklist for observational studies.

Ethical Considerations

Approval and registration

• Approval from the Institutional Ethics Committee is taken before any patient is enrolled. The study follows the ICMR National Ethical Guidelines for Biomedical and Health Research Involving Human Participants (2017).
• The study is registered with the Clinical Trials Registry – India (CTRI) before enrolment if the institution or IEC requires registration of observational studies.

Consent

• Written informed consent is taken in a language the participant understands (English, Hindi or the regional language), using an approved Participant Information Sheet.
• Elderly patients with delirium, encephalopathy or critical illness may lack capacity. For them, consent is taken from the legally acceptable representative, and the patient is re-consented when capacity returns. A patient who declines on recovery has their data withdrawn.
• An impartial witness is present for illiterate participants.

Vulnerability and risk

• Elderly and critically ill patients are a vulnerable group. Participation is voluntary, and refusal does not affect treatment in any way.
• The study is non-interventional. There is no extra blood sampling, procedure or cost to the patient. All investigations and treatment are standard care decided by the treating unit.
• Risk to participants is minimal. It is limited to a possible breach of confidentiality, which is reduced by the measures below.

Confidentiality and data protection

• CRFs carry only the study ID. The key linking IDs to names and hospital numbers is held separately by the investigator.
• Electronic data are password-protected, kept on an institutional device and accessible only to the investigator, guide and statistician.
• Personal data are handled in line with the Digital Personal Data Protection Act, 2023.
• Results are published only in aggregate, with no identifiable information.

Conflict of interest and funding

• The investigators declare no conflict of interest.
• The study needs no external funding.

Expected Outcome and Implications

Expected outcome

• A prospective, KDIGO-based description of AKI in patients aged ≥60 years in this institution, covering:
• the share of community- and hospital-acquired AKI;
• the leading aetiologies and how often AKI is multifactorial;
• the distribution of maximum KDIGO stage;
• nephrotoxic drug exposure and polypharmacy;
• the need for KRT;
• recovery status at discharge, length of stay and in-hospital mortality.
• An estimate of how outcome varies by age group, KDIGO stage, sepsis, frailty and pre-existing CKD.

Implications

• Early recognition: supports a protocol of serial creatinine monitoring and age-appropriate interpretation of creatinine in all elderly admissions.
• Prevention: identifies which preventable causes, such as nephrotoxic drugs, the "triple whammy" combination, contrast exposure and volume depletion, could be targeted by prescribing audits and medication review at admission.
• Prognosis and counselling: gives locally relevant data for discussing prognosis, dialysis decisions and goals of care with elderly patients and their families.
• Resource planning: informs planning of dialysis and ICU capacity for the geriatric medical population.
• Further research: provides a baseline for interventional or quality-improvement studies, such as an AKI electronic alert or a nephrotoxin stewardship programme, and for follow-up studies of AKI-to-CKD transition in the elderly.

Limitations

• Single-centre, tertiary care setting: referral bias towards sicker patients limits how far the results apply to primary or secondary care.
• Baseline creatinine is often unavailable. Using the lowest in-hospital value may under-estimate AKI severity in patients who do not recover, or mislabel unrecognised CKD as AKI.
• Serum creatinine is less sensitive in the elderly because of low muscle mass, so mild AKI may be missed. Newer biomarkers (such as cystatin C or NGAL) are not used.
• The urine output criterion can be applied only in catheterised or ICU patients, so AKI may be under-diagnosed and under-staged in ward patients.
• Aetiology is assigned clinically. Kidney biopsy is rarely performed, so intrinsic renal causes may be misclassified.
• No follow-up after discharge: long-term outcomes such as 90-day recovery, AKD or progression to CKD are not captured.
• Discharge against medical advice, which is common in Indian settings, may bias the recovery and mortality estimates. This is addressed with a sensitivity analysis.
• Observational design: associations with mortality cannot be taken as causal, and the number of deaths limits how many predictors can be adjusted for.
• Consecutive sampling, with screening done mainly on working days, may miss some short-stay patients.

Timeline

Months 1–2: preparation and approval

• Literature search using the listed PubMed strings and finalisation of the review of literature.
• Replacing the assumed mortality (p) with a value from a cited study or departmental records, and recalculating the sample size.
• Finalising the CRF.
• IEC submission and approval; CTRI registration if required.

Month 3: pilot

• Pilot on the first 10–15 eligible patients to test the CRF, screening workflow and the baseline-creatinine rule.
• Minor revisions to the CRF, informing the IEC if needed.
• Recruitment continues, and pilot patients are included if the CRF is unchanged.

Months 3–20: recruitment and follow-up (18 months)

• Enrolment of about 9–10 patients per month.
• Daily follow-up of each patient until discharge or death.
• Monthly data entry and quality check by the guide.
• Interim review at month 11: check the enrolment rate against the target of about 82 patients, and act early if it is behind (for example, by including the Medical ICU census more actively).

Month 21: data cleaning and analysis

• Data cleaning, completion of missing source data, and statistical analysis with the statistician.

Months 22–23: writing

• Writing the results, discussion, summary and conclusions.
• Guide review and revisions.

Month 24: submission

• Final formatting, plagiarism check, binding and submission of the thesis to the university.
• Preparation of a manuscript for journal submission.

Outline of the Case Record Form

Section A: Identification and enrolment

1. Study ID, date of admission, date of enrolment, ward or ICU, and treating unit.
2. Age (years), sex, residence (rural or urban), and socioeconomic status (modified Kuppuswamy scale).
3. Eligibility checklist (inclusion and exclusion criteria, ticked).
4. Consent taken from the patient or the LAR; date of re-consent, if applicable.

Section B: Clinical history

1. Presenting complaints and duration; days of illness before admission.
2. Precipitating events: vomiting or diarrhoea, poor oral intake, fever, recent surgery, urinary symptoms, reduced urine output.
3. Comorbidities: diabetes mellitus, hypertension, coronary artery disease, heart failure, known CKD (stage), chronic liver disease, malignancy, cerebrovascular disease, COPD, benign prostatic hyperplasia.
4. Age-adjusted Charlson Comorbidity Index score.
5. Clinical Frailty Scale score (status 2 weeks before admission).
6. Drug history over the past 7 days, including total number of regular medicines (polypharmacy yes/no). Specific drugs recorded: NSAIDs, ACE inhibitor/ARB, diuretics, SGLT2 inhibitors, metformin, aminoglycosides, vancomycin, other antibiotics, and indigenous or herbal medicines.
7. Iodinated contrast exposure in the past 72 hours (date, procedure).

Section C: Examination at enrolment

1. Pulse, blood pressure, mean arterial pressure, respiratory rate, temperature, SpO₂, and Glasgow Coma Scale.
2. Weight (measured or estimated) and volume status (dehydrated, euvolaemic or overloaded).
3. Systemic examination findings, including bladder palpation, pedal oedema and lung crepitations.
4. SOFA score; APACHE II for ICU patients.

Section D: Investigations

1. Baseline serum creatinine: value, date and source (outpatient value 7–365 days before admission, or lowest in-hospital value).
2. Admission serum creatinine and urea; eGFR (CKD-EPI 2021).
3. Serum sodium, potassium, bicarbonate, calcium, phosphate and uric acid; lactate.
4. Haemoglobin, total leucocyte count and platelets; liver function tests; blood glucose and HbA1c.
5. Urine routine and microscopy, protein-to-creatinine ratio, FENa and FEUrea.
6. Ultrasound KUB findings.
7. Infection and other work-up as indicated: cultures, tropical fever tests, CK, echocardiography, serology, biopsy.

Section E: AKI characterisation

1. Date of AKI diagnosis; community- or hospital-acquired.
2. Criterion met (creatinine, urine output, or both); oliguria (yes/no).
3. KDIGO stage at diagnosis; maximum KDIGO stage and the date it was reached.
4. AKI-on-CKD (yes/no).
5. Aetiology: all contributing causes ticked, with the predominant cause marked.

Section F: Daily hospital course (one row per follow-up day)

1. Day number, serum creatinine, potassium and bicarbonate.
2. Urine output (if charted) and fluid balance.
3. KDIGO stage on that day.
4. Complications: hyperkalaemia, acidosis, pulmonary oedema, encephalopathy, new infection.
5. ICU transfer, mechanical ventilation, vasopressors.
6. Nephrotoxic drugs stopped or dose-adjusted.

Section G: Kidney replacement therapy

1. KRT started (yes/no); date; days from AKI diagnosis.
2. Indication: hyperkalaemia, acidosis, fluid overload, uraemic complication, or progressive azotaemia.
3. Modality (intermittent haemodialysis, SLED or peritoneal dialysis) and number of sessions.
4. KRT declined by the family despite an indication (yes/no).

Section H: Outcome

1. Final status: discharged, died, LAMA, or transferred out.
2. Serum creatinine at discharge; recovery status (complete, partial, or dialysis-dependent).
3. Total length of stay; days from AKI diagnosis to discharge or death; length of ICU stay.
4. If died: date and primary cause of death.

Section I: Verification

1. Investigator's signature and date.
2. Guide's verification signature (for the 10% audit sample).