Perioperative assessment and management in older people

Age alone is not a contraindication. Patients over 70 have higher rates of major perioperative complications and mortality after major non-cardiac surgery, but mortality remains low even beyond 80. Outcome is determined far more by the presence and severity of systemic illness and whether surgery is elective or emergency than by chronological age (Essentials, Table 3-18).

Emergency surgery carries a much higher risk than the same elective operation; wherever possible, convert an emergency into a semi-elective, optimised procedure.

The first question is not "can he survive it?" but "should it be done?" The appropriateness and risk–benefit ratio of the proposed operation must be considered explicitly, in the light of the patient's goals and life expectancy.

Optimise underlying conditions before non-emergency surgery (Essentials, Table 3-18): cardiovascular disease especially heart failure; pulmonary status; renal function; diabetes mellitus; thyroid disease, which is often occult; anaemia; nutrition; and hydration/volume status, particularly in patients on diuretics.

Identify high-risk cardiac conditions. Substantially increased risk is conferred by myocardial infarction within 6 months, pulmonary oedema, angina (especially unstable) and severe aortic stenosis.

Estimate cardiac risk formally. Factors predicting perioperative cardiac complications in non-cardiac surgery are ischaemic heart disease, congestive heart failure, diabetes mellitus, cerebrovascular disease and renal insufficiency (Revised Cardiac Risk Index, Lee et al.).

Assess functional capacity. Exercise tolerance is more informative than laboratory testing — classically, the ability to climb one flight of stairs.

Investigate selectively. Order tests only if the result will change management. Preoperative pulmonary function tests and arterial blood gases are rarely of prognostic value; reserve them for respiratory symptoms, obesity, chest deformity (kyphoscoliosis), abnormal chest radiograph, or planned thoracic/upper abdominal surgery. Measure creatinine clearance where renal function is borderline or nephrotoxic/renally excreted drugs are used.

Assess frailty — the geriatric addition to standard risk scoring. Frailty (unintentional weight loss, weak grip, exhaustion, low physical activity, slow gait) independently predicts postoperative complications, length of stay and discharge to a care facility (Makary et al.). Frailty confers a two- to fourfold increase in operative mortality in cardiac surgery (Pathy, Ch. 35), and markers include Katz ADL ≤4/6, 5-metre walk ≥6 s, 6-minute walk <150 m, and Timed Up-and-Go ≥15 s.

Predict 6-month mortality. Preoperative impaired cognition, recent falls, low albumin, greater anaemia, functional dependence and increased comorbidity are associated with higher mortality at 6 months (Robinson et al.).

Predict delirium. Four independent preoperative predictors: prior stroke or TIA, low MMSE score, abnormal serum albumin, and a high Geriatric Depression Scale score (Rudolph et al.).

Perform a comprehensive geriatric assessment where time allows: medication review, functional assessment, cognition, nutrition, social and caregiver support, and discussion of advance directives before the patient goes to theatre.

Plan the medication regimen deliberately. Some drugs must be continued, some withheld, and some dose-adjusted. Review against Beers/STOPP; inappropriate medication use in older surgical patients is common.

Prophylaxis. Give venous thromboembolism prophylaxis and appropriate antimicrobial prophylaxis according to procedure-specific protocols.

Beta-blockade (textbook position). Patients at high or intermediate cardiac risk with negative non-invasive tests should receive perioperative beta-blockade; those at low risk need none; positive non-invasive tests should prompt consideration of revascularisation. See the Current Practice Update — this position has changed.

Revascularisation is not a licence for surgery. Coronary artery bypass grafting or PCI should be limited to patients with an indication independent of the need for non-cardiac surgery.

Anaesthetic technique should be individualised. Regional anaesthesia is often favoured, but may require supplementary sedation and can itself cause significant cardiovascular change; neither DVT incidence nor blood loss is substantially reduced compared with general anaesthesia. Decide on patient factors, the procedure and the surgical team's preference.

Postoperative care is geriatric care. Prevent delirium (orientation, sensory aids, sleep hygiene, hydration, early mobilisation, avoid restraints and unnecessary catheters), give adequate analgesia, avoid bed rest, attend to nutrition and bowels, and start rehabilitation early. Functional decline in hospital may be irreversible — one-third of older inpatients lose at least one ADL by discharge.

Numbers to quote Item Figure Polypharmacy definitions in the literature ~24 different definitions Polypharmacy / hyperpolypharmacy cut-offs ≥5 drugs / ≥10 drugs Scotland trend, 1995→2010 ≥5 drugs 11%→20%; ≥10 drugs 1.5%→6%; serious interaction exposure 13% (2010) Sweden (1992→2002); Italy (2000→2010) 18%→42%; 43%→53% Medication errors, UK nursing home residents 69.5% incidence Beers Criteria 2019 / 2023 88 PIMs (2019); 5 categories retained in 2023; >3 dozen drugs to avoid, 40+ disease-specific STOPP/START v2 (2015) / v3 (2023) 114 (80+34) / 190 (133+57) — a 66.7% increase Medication Appropriateness Index 10 items, 3-point Likert, maximum score 18 Pill Pruner 13 drug groups, pocket card, effects sustained at 90 days Bisphosphonate holiday Reassess at 5 y oral / 3 y IV; holiday 2–3 y (low risk); continue 10 y oral / 6 y IV (high risk); AFF risk falls ~72%/year after stopping Frailty and operative mortality 2- to 4-fold increase; Katz ≤4/6, TUG ≥15 s, 5-m walk ≥6 s, 6-min walk <150 m Cardiac surgery mortality, age >80 CABG 5–9%; isolated valve 5–7%; combined 8–11%; 5-year survival ≥75 y ≈ 80% Functional decline in hospital One-third of older inpatients lose ≥1 ADL by discharge Multicomponent delirium prevention RR 0.71 (95% CI 0.59–0.85) — 29% relative risk reduction Deprescribing meta-analyses 32 RCTs / 18,670 patients (JAMDA 2023); 259 studies, no mortality benefit (BJCP 2024) Pharmacist-led deprescribing CMR 2.5 vs 1.7 medications stopped/reduced per patient at 6 months; no HRQoL change SGLT2 inhibitors before surgery Stop 3–4 days preoperatively Elective surgery after stroke/TIA Delay ≥3 months Postoperative delirium screening All patients ≥60 years; validated tool daily for ≥3 days from recovery 2. Named criteria, tools and mnemonics Name Content Beers (1991, AGS) 5 categories + anticholinergic table; 2023 update; 2025 Alternatives companion STOPP / START (2008, Europe) PIMs / prescribing omissions; v3 2023, 190 criteria, class-based STOPPFrail For frail patients with limited life expectancy MAI (1992, USA) 10 implicit criteria per drug DBI (2007) Cumulative anticholinergic + sedative exposure FORTA (2008, Germany) A–D labels; first tool to combine positive and negative labelling TRIM (2016, US VA) EHR algorithm using Beers + STOPP SAIL / TIDE Simple, Adverse effects, Indication, List / Time, Individual response, Drug interactions, Educate ARMOR Assess, Review, Minimise, Optimise, Reassess ERASE Resolved conditions, Ageing normally, Selecting targets → Eliminate unnecessary diagnoses AVOID TOO MANY 12-item medication review checklist Scott 5-step List → assess harm → discontinuation potential → prioritise → monitor Barnett 7-step / Deprescribing Goals-first process / clinical, psychological, social, financial and physical determinants Rainbow RCRI (Lee) IHD, CHF, diabetes, cerebrovascular disease, renal insufficiency, high-risk surgery EuroSCORE II / STS score Cardiac surgical operative mortality prediction Rudolph delirium rule Prior stroke/TIA, MMSE, abnormal albumin, GDS Clinical Frailty Scale Recommended preoperatively by ESAIC 2025 (Grade 1C) DASI / METs Functional capacity assessment, 2024 AHA/ACC guideline

Guidelines and evidence, 2023–2026 Source Key message AGS Beers Criteria, JAGS 2023 Aspirin for primary prevention → avoid; warfarin and rivaroxaban → avoid as initial therapy; doxepin removed; SGLT2i caution AGS Beers Alternatives, JAGS 2025 Safer pharmacological and non-pharmacological alternatives across 21 conditions STOPP/START v3, Eur Geriatr Med 2023 190 criteria; class-based; 41–42% new criteria SENATOR / OPERAM / OPTICA No significant reduction in ADRs, mortality, readmissions or QoL JAMDA 2023; BJCP 2024; Age Ageing Deprescribing is safe and reduces PIMs; no mortality or QoL benefit demonstrated 2026 2024 AHA/ACC multisociety Stepwise algorithm; targeted testing; statins Class 1; delay elective surgery ≥3 months perioperative guideline after stroke; OSA screening; MINS recognised; stop SGLT2i 3–4 days ESAIC preoperative assessment, EJA 2025 Frailty testing recommended (1C); Clinical Frailty Scale; WHODAS 2.0 in high-risk patients ESAIC postoperative delirium, EJA 2024 13 recommendations, 5 strong; screen ≥60 y; multicomponent non-pharmacological bundle; EEG depth monitoring; dexmedetomidine risk–benefit; haloperidol only if non-pharmacological measures fail ASA Practice Advisory, Anesthesiology Expanded preoperative evaluation of older inpatients; neuraxial and general anaesthesia 2025 equivalent for delirium risk GLP-1 multisociety guidance, 2024 Continue in most patients; 24-hour clear liquid diet if higher risk; supersedes 7-day withholding of weekly agents WHO Medication Without Harm Third Global Patient Safety Challenge; polypharmacy is a named priority action area Five statements that must not be got wrong

Absorption is the pharmacological parameter least affected by ageing; renal excretion is the most predictable change.

A normal serum creatinine does not mean normal renal function in an older person.

A drug holiday is absolutely contraindicated with denosumab because of rebound vertebral fractures; the levodopa drug holiday is an abandoned, dangerous practice.

Do not start beta-blockers immediately before surgery — continue them if already prescribed. This reverses the older textbook recommendation.

Deprescribing is safe and reduces inappropriate medication use, but no mortality benefit has been demonstrated — state this rather than overclaiming. Textbook content drawn from Pathy's Principles and Practice of Geriatric Medicine (6th ed.) and Essentials of Clinical Geriatrics (7th ed.). Updates cited inline by journal or guideline and year, current to 2026.