Evidence-based nursing practice

In short

Evidence-based nursing practice (EBP/EBN) is the conscientious use of the current best research evidence, combined with the nurse's clinical expertise and the patient's values, preferences and circumstances, to make care decisions — Sackett's three-circle definition, with "available resources" added as a fourth element in the Joanna Briggs model. It replaces practice based on tradition, seniority opinion and "we have always done it this way" with practice that can be shown to work. It is a compulsory unit in B.Sc Nursing under the Indian Nursing Council curriculum (Nursing Research and Statistics, Nursing Management, and the Professional Practice/Introduction to Nursing papers), and appears in M.Sc entrance tests, NORCET/AIIMS and state staff-nurse recruitment, and in NCLEX-RN and OSCE-style viva. On the job it is the basis of hospital nursing protocols, care bundles, NABH-accredited policy manuals, infection-control audits and quality-improvement projects that staff nurses are expected to run.

Explained

What EBP is, and what it is not

• Definition (Sackett, 1996 — the three-component wording is from Sackett et al., 2000): EBM is "the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients" (1996), later stated as "the integration of best research evidence with clinical expertise and patient values" (2000). Research alone is not EBP; expertise alone is not EBP.
• Four components in nursing practice: (1) best available external evidence, (2) clinical judgement and skill, (3) patient preferences, culture and consent, (4) context and resources (staffing, equipment, affordability).
• Related but different terms: *research* generates new knowledge; *research utilization* applies the findings of a single study; *EBP* appraises and synthesises all relevant evidence; *quality improvement (QI)* improves a local process using local data; *evidence-based medicine* is the parent movement (Archie Cochrane's critique of unevaluated practice, 1972).
• Not EBP: "cookbook" care that ignores the individual, cost-cutting dressed as evidence, or a single RCT generalised to every patient.
• Why it matters: reduces practice variation, avoids harmful rituals, shortens stay, lowers hospital-acquired infection and cost, and gives the nurse a defensible basis for action if care is questioned in an inquiry or court.

The five (or seven) steps of EBP

1. Step 0 — Cultivate a spirit of inquiry: notice that a practice may be unsupported (e.g. routine 2-hourly turning for every patient regardless of risk).
2. Step 1 — Ask an answerable clinical question in PICO(T) form.
3. Step 2 — Search for the best evidence in an order: pre-appraised sources first (clinical guidelines, Cochrane Library, JBI, UpToDate, BMJ Best Practice), then primary databases (PubMed/MEDLINE, CINAHL, Embase, Scopus, IndMED, Google Scholar, ERMED and the National Medical Library's e-resources available to Indian institutions).
4. Step 3 — Critically appraise the evidence for validity, reliability, size of effect and applicability — using a checklist (CASP, JBI, PRISMA for reviews, CONSORT for RCTs, STROBE for observational studies).
5. Step 4 — Integrate the evidence with clinical expertise and patient preference and make a decision/protocol.
6. Step 5 — Evaluate the outcome of the change (audit, indicators, patient outcomes).
7. Step 6 — Disseminate the result (ward meeting, policy manual, conference, journal, NABH quality committee).

Framing the question: PICO(T) and question types

• P = population/problem, I = intervention/exposure, C = comparison, O = outcome, T = time frame (optional), and sometimes S = study design.
• *Example:* In post-operative abdominal surgery adults (P), does chlorhexidine-alcohol skin preparation (I) compared with povidone-iodine (C) reduce surgical site infection (O) within 30 days (T)?
• Variants: PICo (Population, Interest, Context) for qualitative questions; SPIDER; PEO for exposure questions.
• Question types and the design that best answers each: therapy/intervention → RCT or systematic review of RCTs; diagnosis/screening → cross-sectional study against a gold standard; prognosis → cohort study; aetiology/harm → cohort or case-control; meaning/experience → qualitative (phenomenology, grounded theory); cost → economic evaluation.
• Background vs foreground questions: "What is sepsis?" is a background question (textbook); "Does early mobilisation reduce ICU-acquired weakness?" is a foreground question (EBP).

Hierarchy and grading of evidence

• Classic pyramid (strongest to weakest): systematic review/meta-analysis of RCTs → single RCT → controlled trial without randomisation → cohort study → case-control study → cross-sectional/descriptive and qualitative studies → case reports → expert opinion, editorials and tradition.
• Melnyk and Fineout-Overholt levels I-VII are the version most Indian nursing syllabi use: Level I systematic review/meta-analysis of RCTs, II single RCT, III controlled trial without randomisation, IV case-control/cohort, V systematic review of descriptive/qualitative studies, VI single descriptive or qualitative study, VII expert opinion.
• JBI levels run 1-5 (effectiveness) and, importantly, add a FAME scale — Feasibility, Appropriateness, Meaningfulness, Effectiveness — recognising that qualitative evidence is the *highest* evidence for questions about experience.
• GRADE rates the *body* of evidence as high, moderate, low or very low certainty, downgrading for risk of bias, inconsistency, indirectness, imprecision and publication bias, and upgrading for large effect or dose-response. Recommendations are strong or conditional (weak).
• Key point: the hierarchy ranks evidence *for effectiveness questions only*. A badly conducted RCT is weaker than a well-conducted cohort study; design is a starting assumption, not a verdict.

Critical appraisal: what to look for

• Internal validity (RCT): true randomisation, allocation concealment, blinding, baseline comparability, intention-to-treat analysis, loss to follow-up under about 20%, adequate power and pre-stated sample size.
• Threats and biases: selection bias, performance bias, detection bias, attrition bias, recall bias (case-control), Hawthorne effect, confounding, publication bias (funnel plot asymmetry).
• Size and precision of effect: read RR, OR, ARR, NNT, mean difference and the 95% confidence interval. A CI that crosses 1 for RR/OR (or 0 for a difference) means the result is not statistically significant. Statistical significance (p < 0.05) is not clinical significance.
• Heterogeneity in meta-analysis: I² roughly >50% is substantial; forest plot shows each study's effect and the pooled diamond.
• Reliability and validity of tools: Cronbach's alpha ≥0.7 acceptable for internal consistency; content validity index; test-retest reliability; inter-rater kappa.
• Applicability to India: were the patients, nurse-patient ratio, equipment and drug availability comparable? A trial run with 1:1 ICU nursing may not transfer to a ward with 1:20.
• Qualitative appraisal: credibility, transferability, dependability, confirmability (Lincoln and Guba), reflexivity, triangulation, member checking, data saturation.

Models of EBP implementation

• Stetler model — five phases: preparation, validation, comparative evaluation/decision-making, translation/application, evaluation; practitioner-oriented.
• Iowa model — begins with a problem-focused or knowledge-focused *trigger*, asks "is this a priority for the organisation?", forms a team, pilots the change, then adopts or rejects it.
• ACE Star model (Stevens) — five points: discovery of research, evidence summary, translation into guidelines, integration into practice, evaluation.
• Rosswurm and Larrabee model — assess need, link problem with interventions and outcomes, synthesise best evidence, design change, implement and evaluate, integrate and maintain.
• PARIHS/i-PARIHS framework — successful implementation = f(Evidence × Context × Facilitation); explains why good evidence fails in a hostile ward culture.
• Knowledge-to-Action (Graham) and Ottawa Model of Research Use — cyclical adaptation of knowledge to local context.
• Stages in a hospital: evidence → clinical practice guideline → local protocol/SOP/standing order → care bundle and checklist → audit indicator.

Barriers, facilitators and the Indian context

• Barriers reported in Indian studies: heavy workload and poor nurse-patient ratio, no protected time, limited access to paid journals, English-language and statistics difficulty, lack of library/internet at the bedside, doctor-dominated decision-making, resistance from senior staff, no institutional reward, and unavailability of the recommended consumable.
• Facilitators: nursing research committee, journal club, EBP mentors/champions, nursing informatics and subscribed databases, in-service education, protected study time, support from the nursing superintendent, and accreditation pressure.
• Indian structures that drive EBP:
  • Indian Nursing Council curricula require research, statistics and EBP; the National Nursing and Midwifery Commission Act, 2023 has been enacted to replace the Indian Nursing Council Act, 1947 — the Commission's constitution and regulations are being rolled out, so cite the Act, and check the current regulator's notifications before quoting detailed rules.
  • NABH accreditation standards require documented, evidence-based clinical policies, care protocols, nursing care plans and clinical audits; NQAS (National Quality Assurance Standards) and Kayakalp do the same for public facilities under the National Health Mission.
  • ICMR National Ethical Guidelines for Biomedical and Health Research Involving Human Participants (2017), Institutional Ethics Committee clearance, and the New Drugs and Clinical Trials Rules, 2019 govern nursing research; CTRI (Clinical Trials Registry – India) registers trials.
  • National guidelines nurses apply daily: NCDC/NCVBDC programme guidelines, NACO HIV guidelines, RNTCP/NTEP TB treatment guidelines, IMNCI/NSSK newborn care, LaQshya labour-room standards, MusQan, AMR/HIC guidelines of NCDC and the ICMR antimicrobial stewardship programme, and the national Standard Treatment Guidelines/STGs used in Ayushman Bharat PM-JAY.
  • Nursing-specific sources: JBI, Cochrane Nursing Care Field, Registered Nurses' Association of Ontario (RNAO) BPGs, WHO guidelines, CDC/WHO hand-hygiene and CLABSI/CAUTI/VAP bundle guidance, NPIAP/EPUAP pressure-injury guideline, Braden and Morse risk scales.
• Examples of evidence changing Indian nursing practice: WHO's "My Five Moments" hand hygiene; abandoning routine pre-operative shaving with a razor in favour of clippers or no hair removal; kangaroo mother care and delayed cord clamping; early enteral feeding; no routine enema or episiotomy in normal labour; saline flush instead of heparin for peripheral cannula; aseptic non-touch technique; no routine IV site rotation at fixed hours in favour of clinically indicated replacement; chlorhexidine oral care and head-of-bed elevation 30-45° for VAP prevention (chlorhexidine body bathing is for CLABSI/MRSA, not VAP).

Key points

• EBP - best research evidence + clinical expertise + patient values + context/resources.
• Sackett - author of the working definition; Archie Cochrane - inspired the movement; Florence Nightingale - first nurse to use statistics to change practice.
• Five steps - ask (PICO), search, appraise, apply/integrate, evaluate (plus step 0 inquiry and step 6 dissemination).
• PICOT - Population, Intervention, Comparison, Outcome, Time; converts a vague concern into a searchable question.
• Level I evidence - systematic review or meta-analysis of RCTs; Level VII - expert opinion.
• GRADE - grades certainty of a body of evidence (high to very low) and strength of recommendation (strong/conditional).
• JBI FAME - Feasibility, Appropriateness, Meaningfulness, Effectiveness; qualitative evidence is highest for experience questions.
• Critical appraisal tools - CASP, JBI checklists, PRISMA (reviews), CONSORT (RCTs), STROBE (observational), COREQ (qualitative).
• RR/OR < 1 - protective; = 1 - no effect; > 1 - increased risk; significance judged by the 95% CI crossing 1.
• NNT - number needed to treat = 1/ARR; smaller is better; NNH is its harm counterpart.
• Implementation models - Stetler, Iowa, ACE Star, Rosswurm and Larrabee, PARIHS, Knowledge-to-Action.
• PARIHS - implementation success depends on evidence × context × facilitation, not evidence alone.
• Guideline vs protocol vs bundle - guideline is national and advisory; protocol/SOP is local and binding; a bundle is 3-5 evidence-based elements done together, all-or-none.
• Nursing audit - measures whether practice matches the standard; the evaluation arm of EBP (structure-process-outcome, Donabedian).
• Research utilization - applying one study's finding; narrower than EBP.
• Indian regulators and standards - INC curriculum, National Nursing and Midwifery Commission Act 2023, NABH, NQAS, ICMR 2017 ethics guidelines, CTRI registration.
• Biggest Indian barriers - workload, no protected time, journal access, statistics confidence; champions and journal clubs are the common remedies.
• Clinical significance ≠ statistical significance - a tiny effect can be significant in a huge sample and useless at the bedside.

Formulas and examples

Example 1 — ARR, RRR and NNT for a pressure-injury bundle. In a 400-bed Indian tertiary hospital, 200 high-risk patients receive usual care and 200 receive a bundle (Braden scoring, 4-hourly repositioning, foam dressing on the sacrum, nutrition referral). Pressure injuries: 60 in usual care, 36 in the bundle group.

• CER = 60/200 = 0.30 (30%); EER = 36/200 = 0.18 (18%).
• RR = 0.18/0.30 = 0.60 → the bundle leaves 60% of the risk, i.e. RRR = 1 − 0.60 = 40%.
• ARR = 0.30 − 0.18 = 0.12 = 12 percentage points.
• NNT = 1/0.12 = 8.33 → 9 patients must receive the bundle to prevent one pressure injury.
• Cost check: bundle consumables ₹250 per patient → 9 × ₹250 = ₹2,250 to prevent one injury, against a conservative ₹40,000 to treat one stage-3 ulcer (dressings, extra stay, antibiotics) — strongly cost-saving.
• 95% CI for the bundle group's 18%: SE = √(0.18 × 0.82 / 200) = √0.000738 = 0.0272 → 0.18 ± 1.96 × 0.0272 = 0.18 ± 0.0533 → 12.7% to 23.3%, which does not overlap 30%, so the difference is not likely to be chance.

Example 2 — odds ratio from a case-control study on inpatient falls. 100 patients who fell (cases) and 100 matched non-fallers (controls) are reviewed for sedative use at night: cases exposed a = 40, cases unexposed b = 60; controls exposed c = 20, controls unexposed d = 80.

• OR = (40 × 80)/(60 × 20) = 3,200/1,200 = 2.67.
• Interpretation: the odds of night sedation are about 2.7 times higher among fallers; this is an *association* from a retrospective design (Level IV), so it supports a medication-review protocol, not a causal claim.

Example 3 — appraising a delirium screening tool. 400 ICU patients are screened with a nurse-administered tool and compared with a psychiatrist's diagnosis (gold standard): TP = 90, FN = 10, FP = 30, TN = 270.

• Prevalence = (90 + 10)/400 = 25%.
• Sensitivity = 90/100 = 90%; Specificity = 270/300 = 90%.
• PPV = 90/(90 + 30) = 90/120 = 75%; NPV = 270/(270 + 10) = 270/280 = 96.4%.
• LR+ = 0.90/(1 − 0.90) = 9.0; LR− = (1 − 0.90)/0.90 = 0.11.
• Conclusion: good rule-out tool (NPV 96.4%); a positive screen needs confirmation because 1 in 4 positives is false. Note that PPV falls if the same tool is used on a low-prevalence general ward.

Example 4 — evaluating the change (step 5) with audit indicators. Before the CAUTI bundle: 8 infections in 1,250 catheter-days. After: 3 infections in 1,100 catheter-days.

• Before = (8/1,250) × 1,000 = 6.4 per 1,000 catheter-days; After = (3/1,100) × 1,000 = 2.73 per 1,000 catheter-days.
• Relative reduction = (6.4 − 2.73)/6.4 = 3.67/6.4 = 57.3%.
• Hand-hygiene compliance audit in the same quarter: 168 of 240 opportunities before = 70%; 204 of 240 after = 85%.

Where it is used

• Hospital nursing administration and quality cells: writing and revising policy/procedure manuals, care plans and SOPs for NABH or JCI accreditation; roles titled Quality Nurse, Nursing Quality Coordinator, NABH Coordinator in Apollo, Fortis, Manipal, Max, Narayana Health, AIIMS, PGIMER, CMC Vellore, Aster and Amrita.
• Infection prevention and control: Infection Control Nurse running surveillance per NCDC/ICMR and CDC-NHSN definitions, hand-hygiene audits, CLABSI/CAUTI/VAP/SSI bundles, antimicrobial stewardship support.
• Clinical research and trials: Clinical Research Coordinator/Research Nurse posts in CROs and hospital research units, working to ICMR 2017 ethics guidelines, New Drugs and Clinical Trials Rules 2019, ICH-GCP and CTRI registration; also ICMR/DBT-funded nursing research at colleges.
• Nursing education: faculty teaching Nursing Research and Statistics, running journal clubs, guiding M.Sc dissertations, developing and validating tools and modules.
• Public health and programme nursing: community health officers and PHNs applying IMNCI, NSSK, LaQshya, NTEP, NACO and immunisation guidelines at Ayushman Arogya Mandirs under NQAS/Kayakalp assessment.
• Clinical specialist roles: ICU, dialysis, oncology, neonatal and wound-care nurses selecting risk scales (Braden, Morse, NEWS, Glasgow), early-warning systems and protocol-based nursing interventions; NABH-driven clinical audits and incident/root-cause analysis.

Interview questions

• Q: Define evidence-based nursing practice. A: The integration of the current best research evidence with the nurse's clinical expertise and the patient's values, preferences and circumstances — and, in the JBI formulation, the available context and resources — when making care decisions.
• Q: What are the steps of EBP? A: Cultivate inquiry, ask a focused PICO question, search for the best evidence, critically appraise it, integrate it with expertise and patient values, evaluate the outcome, and disseminate the result.
• Q: Write a PICO question for a problem on your ward. A: For example: in adult patients with peripheral IV cannulas (P), does clinically indicated replacement (I) compared with routine 72-hourly replacement (C) change the rate of phlebitis (O) during admission (T)?
• Q: Which is the highest level of evidence, and why not simply follow it always? A: A systematic review or meta-analysis of RCTs; it must still be checked for quality, heterogeneity and applicability — the trial population, staffing and resources may not match your setting.
• Q: Difference between research, research utilization and EBP? A: Research creates new knowledge; research utilization applies the findings of a particular study; EBP appraises and synthesises all the relevant evidence along with clinical judgement and patient values.
• Q: A drug reduces mortality from 4% to 3% — explain in terms a ward sister can use. A: ARR is 1 percentage point and RR is 0.75, so RRR is 25%; NNT = 1/0.01 = 100, meaning 100 patients must be treated to prevent one death — the 25% figure alone sounds much larger than the real benefit.
• Q: What does a 95% confidence interval for RR of 0.80 (0.62 to 1.03) tell you? A: The point estimate suggests a 20% risk reduction, but because the interval includes 1 the result is not statistically significant and no benefit can be claimed.
• Q: How is qualitative research useful if it is low in the evidence hierarchy? A: The hierarchy applies to effectiveness questions; for feasibility, appropriateness and meaning — why patients refuse an intervention, for instance — well-conducted qualitative research is the highest-quality evidence, as JBI's FAME scale recognises.
• Q: Name the EBP models you know and one feature of each. A: Iowa (starts from a problem- or knowledge-focused trigger and pilots the change), Stetler (practitioner-focused, five phases), ACE Star (five points from discovery to evaluation), Rosswurm and Larrabee (six-step change model), PARIHS (evidence × context × facilitation).
• Q: What are the main barriers to EBP in Indian hospitals, and what would you do about one? A: Workload and nurse-patient ratio, no protected time, limited database access and statistics confidence; I would start a monthly journal club with one appraised article translated into a single protocol change, supported by the nursing superintendent.
• Q: Which appraisal checklist would you use for a cohort study and for a systematic review? A: STROBE or the CASP/JBI cohort checklist for the cohort study, and PRISMA with the CASP/JBI systematic-review checklist for the review.
• Q: What Indian requirements apply if you want to conduct a nursing study in your hospital? A: Institutional Ethics Committee approval under the ICMR National Ethical Guidelines (2017), written informed consent, confidentiality and data protection, and — for a clinical trial — CTRI registration and compliance with the New Drugs and Clinical Trials Rules, 2019.
• Q: If the best evidence conflicts with the patient's wishes, what do you do? A: Explain the evidence and its uncertainty, respect the competent patient's informed refusal, document it, and offer the best acceptable alternative — autonomy is one of the three pillars of EBP, not an exception to it.

Common mistakes

• Defining EBP as "following research findings" and omitting clinical expertise, patient values and context — the most commonly lost marks in theory papers.
• Reporting only RRR (sounds impressive) without ARR and NNT, or quoting NNT as a decimal instead of rounding up to a whole patient.
• Treating a p value as proof of clinical importance, and ignoring the confidence interval and effect size.
• Confusing sensitivity with PPV, and forgetting that PPV and NPV change with prevalence while sensitivity and specificity do not.
• Calculating an odds ratio from a cohort or RCT when risk ratio is appropriate, or swapping the a·d and b·c products.
• Assuming the highest-level study automatically answers the question — applying an effectiveness hierarchy to a question about patient experience, or ignoring that the trial setting is unlike an Indian ward.
• Stopping at Step 4: implementing a change and never auditing the outcome, so the EBP cycle is never closed and the practice is not sustained.

Related topics

Nursing research and statistics · Nursing ethics, documentation and law · Nursing management and leadership · Quality assurance, nursing audit and NABH standards · Infection prevention and control · Nursing theories and the nursing process