Quality Improvement, Registries & IBD Research
Module Overview
IBD registry design and participation, quality metrics, clinical trial methodology, outcomes research, and quality improvement projects.
Benchmark Source: Cleveland Clinic Research Track + IOIBD Research Curriculum
Learning Objectives
Design and contribute to an IBD patient registry
Apply IBD quality metrics (ICHOM, AGA Quality Measures)
Critically appraise IBD clinical trial methodology
Develop a quality improvement project proposal
Teaching Sessions
IBD Registries: Design, Ethics & Saudi Context
Registry design, REDCap implementation, ethics/PDPL compliance, and participation in OASIS/AMBER Saudi registries.
1What is a Registry? Types, Purposes, Value
A patient registry is an organized system that uses observational methods to collect uniform clinical and other data to evaluate specified outcomes for a population defined by a particular disease, condition, or exposure. Types: (1) disease registries (all IBD patients); (2) product registries (all patients on a specific biologic—e.g., TREAT for infliximab); (3) health services registries (tracking quality metrics); (4) pregnancy registries (PIANO, SAPPHIRE); (5) device registries. Purposes: (a) natural history studies—describe disease course over time; (b) comparative effectiveness—real-world comparisons of therapies; (c) safety surveillance—detect rare adverse events; (d) quality improvement—benchmark and improve care; (e) health policy—inform resource allocation, formulary decisions. Value over RCTs: external validity (representative populations, real-world use); long-term follow-up (decades); ability to study rare events; cost-effective compared to trials. Limitations: confounding by indication (patients starting a therapy differ from comparators—use propensity scoring, instrumental variables); missing data; inconsistent data quality. Major IBD registries: TREAT (infliximab), PYRAMID (adalimumab), Swiss IBD Cohort Study (SIBDCS), ENEIDA (Spain), EpiCom, CESAME (France), IBD Sinai cohort, ICURE (Canada). Saudi and regional: OASIS (SGA national), AMBER (Middle East observational), SAUDI-IBD (Mosli 2019, KFSHRC-led).
Registries are where real-world evidence lives. A fellow who contributes to one builds a research portfolio that follows them for decades.
- Five registry types: disease, product, services, pregnancy, device
- Registries > RCTs for external validity, long-term follow-up, rare events
- Confounding by indication = use propensity scoring
- Saudi: OASIS (SGA), AMBER (regional), SAUDI-IBD
- Fellow contribution builds long-term research portfolio
2Registry Design: Variables, Case-Report Forms, Data Quality
Design steps: (1) define aims precisely (descriptive? comparative? surveillance?); (2) define eligible population (inclusion/exclusion); (3) specify variables—core dataset essential, optional expanded dataset for sub-studies; follow IBD common data element standards (Fiorino 2020, ICHOM IBD Standard Set, ECCO CONFER); (4) design case-report forms (CRFs)—enrollment, follow-up visits, outcome events; include mandatory fields, validation rules, branching logic; (5) define data dictionary (variable names, definitions, coding, units); (6) standardize definitions (e.g., clinical remission = PRO-2 with RB=0 and SF≤1 in UC, or HBI<5 in CD; biochemical remission = CRP <5 mg/L AND calprotectin <150–250 µg/g; endoscopic remission = Mayo 0–1 or SES-CD <3); (7) data entry training; (8) data quality monitoring—range checks, logic checks, 10% source data verification; (9) adjudication for endpoints (blinded review by committee). Core variables for any IBD registry: demographics, diagnosis date, Montreal phenotype, EIM, family history, smoking, prior surgeries, prior medications, current medication with start/stop dates, disease activity (PRO-2, HBI, SCCAI, UCDAI), CRP, calprotectin, Hb, albumin, endoscopic findings, imaging, complications, hospitalizations, surgeries, pregnancy outcomes, PROs (IBDQ, PRO-2), adverse events. REDCap (Research Electronic Data Capture, Vanderbilt, free for academic use) is the global standard—supports CRFs, branching logic, data quality, audit trails, API access, mobile data entry.
Start with the ICHOM IBD Standard Set—it's free, validated, international, and already has Arabic translations of patient-reported outcomes.
- ICHOM IBD Standard Set = starting point for variables
- REDCap = global standard platform, free for academics
- Standardize definitions (remission, response) up front
- 10% source data verification for quality
- Core + optional datasets support sub-studies
3Ethics, Consent & PDPL Compliance in Saudi Arabia
Ethics framework: all registry research requires IRB approval (institutional review board at each site). Multi-center Saudi studies: central IRB model via SFDA or lead site, with local IRB acknowledgment. Informed consent: registry requires prospective written consent unless IRB-approved waiver (minimal risk + impracticable + no adverse effect on rights); consent should cover: data collected, storage duration, sharing with collaborators, linkage to other datasets, future unspecified research, withdrawal rights, contact information. Saudi-specific consent considerations: bilingual (Arabic + English); family involvement culturally appropriate but patient autonomy primary; literacy assessment and verbal consent with witness if low literacy. Personal Data Protection Law (PDPL, Saudi Arabia, 2023, amended 2024): regulates collection, processing, storage, and transfer of personal data; key requirements for registries: (1) lawful basis (consent or legitimate interest); (2) purpose limitation (specify and limit); (3) data minimization; (4) accuracy; (5) storage limitation; (6) integrity and confidentiality; (7) accountability. Cross-border transfers: restricted unless recipient country has adequate data protection or contractual safeguards; many international IBD registries require Data Transfer Agreements. SDAIA (Saudi Data & AI Authority) oversees PDPL compliance. Practical steps: (1) IRB-approved protocol; (2) Data Privacy Impact Assessment (DPIA) for large registries; (3) pseudonymization (replace identifiers with code), separate keys; (4) encrypted storage, access logs; (5) retention schedule (typically 10–15 years post-study); (6) breach notification within 72 hours. Ramadan/Hajj timing: plan recruitment around cultural calendar.
PDPL is not optional. A registry without pseudonymization, access logs, and a breach plan will fail SDAIA audit—build compliance into design from day one.
- IRB approval + informed consent (bilingual) mandatory
- PDPL 2023/2024: lawful basis, purpose, minimization, accuracy
- Pseudonymization + access logs + breach notification <72h
- Cross-border transfer requires adequacy or DTA
- SDAIA oversight — build compliance from design phase
4OASIS & AMBER: Saudi Participation Pathway
OASIS (Outcomes and Assessment of IBD in Saudi Arabia) is the SGA's national prospective IBD registry—launched to capture real-world outcomes across tertiary centers (KFSHRC, KFMC, KAMC, KAUH, KSUMC, KSMC, and increasingly regional hospitals). Goals: (1) describe Saudi IBD phenotype and natural history (early-onset, male predominance, frequent penetrating Crohn's); (2) benchmark quality of care; (3) support pragmatic comparative effectiveness research; (4) inform health policy and SFDA formulary decisions; (5) training ground for fellows. AMBER (A Middle Eastern Registry) extends this across GCC and Middle East, linking Saudi data to regional comparisons. Fellow participation pathway: (1) complete GCP (Good Clinical Practice) training, REDCap certification, institutional research ethics training; (2) register as site investigator or data entry team; (3) enroll consecutive new IBD patients (with consent); (4) complete enrollment CRF (baseline clinical, phenotype, labs, endoscopy); (5) complete follow-up CRFs at defined intervals (6-month, 12-month, annually); (6) report major events (surgery, hospitalization, pregnancy, death); (7) contribute to sub-studies (pregnancy, pediatric-adult transition, biologic comparative effectiveness); (8) participate in data-sharing proposals and co-author publications. Quality standards: <10% missing core data, <5% discrepancy rate in source verification, adherence to visit windows (±30 days). Outputs: annual SGA scientific report, publications (e.g., Mosli 2023 Saudi IBD Registry), abstracts at DDW, ECCO, regional meetings, PhD/Master thesis projects. For fellows: aim to lead one sub-study or QI analysis from registry data during fellowship year.
Every SGA fellow should exit fellowship with at least one OASIS-linked abstract or publication. It sets up your research identity for the rest of your career.
- OASIS = SGA national registry; AMBER = regional extension
- Fellow pathway: GCP → REDCap → site enrollment → sub-study
- Quality: <10% missing data, <5% discrepancy, visit window ±30d
- Outputs: publications, abstracts, thesis projects
- Career aim: ≥1 OASIS-linked output per fellowship year
- Registries provide real-world evidence, long-term follow-up, rare event detection
- ICHOM IBD Standard Set + REDCap = practical design foundation
- PDPL 2023 mandates consent, pseudonymization, breach plan — not optional
- OASIS/AMBER = fellow research gateway; aim for 1 output per year
Quality Metrics in IBD Care
ICHOM standard set, AGA IBD Quality Measures 2.0, unit-level benchmarking, and fellow-led dashboards.
1Why Measure Quality? Frameworks & Donabedian
Donabedian framework (1966): quality has three domains—structure (resources, staff, facilities), process (what we do—screening, monitoring, treatment delivery), outcomes (patient results—clinical, PROs, mortality). In IBD, all three matter: structure (availability of IBD clinic, MDT, biologic infusion suite, dietitian, psychologist, surgical backup); process (vaccination, TB screening before biologics, biomarker monitoring, dysplasia surveillance); outcomes (steroid-free remission, surgery rates, hospitalization rates, PROs, mortality, cancer incidence). Why measure: (1) patient safety; (2) reduce unwarranted variation; (3) benchmark against peer units and international standards; (4) continuous improvement; (5) regulatory requirements (SFDA, MOH); (6) value-based care as Saudi health system evolves under Vision 2030. Quality vs quantity: measuring what matters to patients (PROs, shared decisions) vs what's easy to measure (visit counts, test ordering). Triple Aim (IHI): better care, better health, lower cost. Quadruple Aim adds clinician well-being. Balancing measures: always track unintended consequences (e.g., cutting colonoscopy costs might reduce surveillance quality). Avoid gaming: well-designed metrics are objective, transparent, risk-adjusted, and trigger learning not punishment.
A metric that cannot be gamed is one where patients would notice the shortcut. Steroid-free remission cannot be faked; appointment counts can.
- Donabedian: structure, process, outcomes
- Quadruple Aim: care, health, cost, clinician well-being
- Balance measures catch unintended consequences
- Good metrics: objective, transparent, risk-adjusted
- Vision 2030 drives value-based IBD care in Saudi
2AGA IBD Quality Measures 2.0 & ICHOM Standard Set
AGA IBD Quality Measures 2.0 (Dulai 2023, Gastroenterology): nine core measures grouped into process and outcomes: PROCESS—(1) corticosteroid-sparing therapy for steroid-dependent patients; (2) tobacco cessation counseling for CD smokers; (3) influenza vaccination annually; (4) pneumococcal vaccination; (5) bone density testing for long-term steroid users; (6) TB screening before immunosuppression; (7) HBV screening before biologics; (8) colorectal cancer surveillance per guidelines. OUTCOME—(9) steroid-free clinical remission at 12 months. Measure specifications: numerator/denominator, exclusions, risk adjustment, data source, reporting interval. Use for practice feedback, MIPS (US Medicare) reporting, CQC (UK) equivalent, Saudi CBAHI accreditation alignment. ICHOM IBD Standard Set (Kim 2022, Lancet Gastroenterol): patient-centered, includes 24 variables across baseline (demographics, phenotype), process (medications, surveillance), and outcomes (PROs including PRO-2, IBDQ-9/32, disability index, work productivity, steroid-free remission, surgery, hospitalization, mortality, treatment burden—infusion visits, side effects). ICHOM includes validated PROs in multiple languages including Arabic. Implementation at unit level: select ~5 priority metrics, automate data capture from EMR where possible, display on team dashboard, monthly review, PDSA cycles for underperformance. Saudi IBD quality scorecard (emerging): SGA is piloting a unified quality dashboard across centers—fellows can contribute by leading local implementation.
Every IBD clinic should track at minimum: steroid-free remission rate, vaccination compliance, surveillance compliance. These three alone drive transformation.
- AGA Quality Measures 2.0 = 9 measures (8 process + 1 outcome)
- ICHOM IBD Standard Set = 24 variables, Arabic validated PROs
- ~5 priority metrics on automated dashboard at unit level
- Monthly review + PDSA for underperformance
- Align with SGA emerging quality scorecard
3Implementing a Personal IBD Dashboard
Practical exercise: each fellow creates a personal IBD dashboard to track performance across a defined patient panel during fellowship. Steps: (1) define panel (patients under your continuity care in the IBD clinic); (2) select metrics—minimum: steroid-free remission at 6 and 12 months, calprotectin in previous 6 months, annual vaccination status, surveillance colonoscopy compliance (Month 3 content), anxiety/depression screening completion, PRO-2 score captured, smoking cessation offered to CD smokers; (3) data source—EMR extract (Excel or REDCap export), supplemented by patient-reported forms; (4) update monthly; (5) compare to unit benchmark and AGA targets (e.g., steroid-free remission >70%, vaccination >90%); (6) identify lowest-performing metric and launch PDSA improvement cycle; (7) share dashboard with program director quarterly. Tools: Excel (simple start), REDCap dashboards, Power BI (hospital-deployed), Tableau. Example PDSA cycle for low vaccination rate: Plan—identify barriers (forgot to order, patient hesitancy, cost); Do—pilot EMR alert for unvaccinated patients + educational card; Study—measure vaccination rate after 3 months; Act—spread or refine based on results. Publish QI outcomes using SQUIRE 2.0 reporting standards. Saudi context: align with SGA quality initiative; publish in Saudi J Gastroenterol or SGA conference. Career value: a fellow who completes one QI project and publishes outcomes is distinguishable on any international fellowship or faculty application.
Start your dashboard in the first month of fellowship with just 3 metrics. Add more as you gain capacity. By the end of the year, you'll have an publishable QI project.
- Personal dashboard: 5–7 metrics on your patient panel
- Monthly update, quarterly review with PD
- PDSA cycle for lowest metric
- SQUIRE 2.0 reporting for publication
- One published QI project = distinguished fellowship output
- Donabedian structure/process/outcomes + Quadruple Aim
- AGA 9 core measures + ICHOM 24-variable standard set
- Build personal dashboard in month 1 of fellowship
- One published QI project = career differentiator
Clinical Trial Design Workshop
Phase 2/3 IBD trial design, endpoints, statistical planning, regulatory pathways (FDA, EMA, SFDA) and Saudi site readiness.
1IBD Trial Architecture: Induction, Maintenance, Treat-Through
Modern IBD trial design has evolved substantially. Induction: typically 6–14 weeks of randomized blinded treatment to demonstrate superiority over placebo in patients with moderate-severe active disease (Mayo 6–12 for UC, CDAI ≥220 or modified SES-CD for CD). Maintenance: responders from induction re-randomized to active vs placebo for 44–52 weeks (randomized withdrawal design) OR continue on induction allocation (treat-through design). Treat-through (e.g., Sequence, Velvet, recent Janssen/AbbVie trials): newer, more pragmatic, avoids "enrichment" bias of responders-only maintenance, gives cleaner effect sizes but requires larger sample. Dose-ranging (Phase 2): determine dose/regimen; Phase 3 confirms efficacy at selected dose. Active comparator trials: directly compare drug vs drug (e.g., SEAVUE compared ustekinumab vs adalimumab in biologic-naïve CD; ADVANCE compared upadacitinib vs adalimumab). Head-to-head data guide sequencing decisions. Open-label extensions: follow responders for years to assess long-term safety/efficacy. Platform trials: rare in IBD but emerging (master protocol, multiple drugs, shared control). Adaptive trials: pre-specified modifications (sample size re-estimation, arm dropping). Subgroup pre-specification: biologic-naïve vs exposed, disease duration, baseline CRP/calprotectin. Quality of life and disability outcomes increasingly central.
The treat-through design is replacing randomized withdrawal because it answers the clinically relevant question: "If I start this drug, what's the chance my patient is in remission at a year?"
- Induction (6–14 wk) → maintenance (44–52 wk) = classic design
- Treat-through = cleaner, more pragmatic, larger sample
- Active comparator trials = sequencing guidance
- Subgroup pre-specification: biologic exposure, CRP, duration
- QoL/disability outcomes increasingly central
2Endpoints: Symptomatic, Endoscopic, Histologic, Biomarker, PRO
FDA/EMA endpoints have evolved from purely symptomatic to composite endpoints. Contemporary UC endpoints: clinical remission (Adapted Mayo: rectal bleeding = 0, stool frequency ≤1 with ≥1-point decrease, endoscopic subscore ≤1); endoscopic improvement (Mayo endoscopic subscore ≤1); endoscopic remission (Mayo 0); histologic remission (Geboes ≤2 or Nancy ≤1); histo-endoscopic mucosal improvement (increasingly used); symptomatic remission (PRO-2: RB=0 + SF reduced). Contemporary CD endpoints: clinical remission (CDAI <150 or PRO-2 CD: unformed stool ≤1.5 and abdominal pain ≤1); endoscopic response (SES-CD ≥50% decrease); endoscopic remission (SES-CD <3 or no ulcerations); transmural healing on MRE (emerging). Steroid-free remission at week 44–52 = key regulatory endpoint for maintenance. Central reading: all endoscopies video-captured and read blindly by central panel—eliminates site bias, standardizes interpretation. PROs: PRO-2 (derived symptomatic subset of validated indices), IBDQ-9/32 (disease-specific QoL), SIBDQ (short form), work productivity (WPAI), disability (IBD-DI). Biomarkers: CRP <5 mg/L, fecal calprotectin <150–250 µg/g—increasingly required as composite or confirmatory endpoints. Safety endpoints: infections (serious, opportunistic, TB reactivation), malignancy, MACE, VTE, injection-site reactions, immunogenicity.
Endoscopic improvement is now required by FDA for UC approval. Central reading is mandatory for registration trials—learn its methodology to succeed as a trial site investigator.
- Composite endpoints: symptomatic + endoscopic + histologic + biomarker
- Central reading = regulatory standard, eliminates site bias
- Steroid-free remission at wk 52 = key maintenance endpoint
- PROs and disability now co-primary or key secondary
- Safety: serious infection, TB, malignancy, MACE, VTE
3Statistical Design: Sample Size, Imputation, Interim Analyses
Sample size calculation requires: (1) primary endpoint and effect size (based on prior trials/meta-analyses); (2) placebo response rate (higher in IBD than most diseases, 10–30% clinical remission on placebo); (3) active drug response rate (40–60% for modern agents); (4) power (typically 80–90%); (5) alpha (0.05 two-sided, adjusted for multiplicity); (6) attrition (20–30% dropout typical). Example: UC Phase 3 induction, 40% active vs 15% placebo remission, 90% power, alpha 0.025 (two comparator arms vs placebo)—approximately 300 per arm. Missing data handling: pre-specified in Statistical Analysis Plan (SAP); most common approach in IBD is Non-Responder Imputation (NRI)—patients with missing endpoint assessment, or who discontinued for lack of efficacy, adverse event, or prohibited medication, classified as non-responders; more conservative than LOCF (last observation carried forward) or MI (multiple imputation). Intention-to-treat (ITT) vs per-protocol (PP): ITT is primary analysis; PP sensitivity. Interim analyses: pre-specified stopping rules for futility (conditional power <20%) or overwhelming efficacy (p < Haybittle-Peto or O'Brien-Fleming boundary); DSMB (Data Safety Monitoring Board) reviews. Multiple comparisons: hierarchical testing (primary → secondary in order), gatekeeping procedures. Subgroup analyses: pre-specified and hypothesis-driven only; interactions tested formally; never data-dredge. Non-inferiority vs superiority: non-inferiority requires pre-specified margin based on clinically important difference (e.g., 10% for UC remission)—margin must be justified and not too generous. Real-world: pragmatic trials with fewer exclusions and real-world outcomes are increasing (e.g., PRECIS-2 framework).
Non-Responder Imputation (NRI) is the standard handling of missing data in IBD trials. If you see a trial using LOCF, read the protocol carefully—NRI is more conservative and less manipulable.
- Sample size needs placebo response (10–30%), effect size, power, attrition
- NRI = standard missing-data handling in IBD
- ITT primary, PP sensitivity
- Hierarchical testing for multiple comparisons
- Pragmatic/PRECIS-2 trials increasingly important
4Regulatory Pathway & Saudi Site Readiness
Regulatory milestones: IND/CTA (Investigational New Drug/Clinical Trial Authorization) to initiate; Phase 1 (healthy volunteers, PK/PD); Phase 2 (dose-ranging, proof of concept in patients); Phase 3 (pivotal efficacy and safety, typically 2 positive trials for registration); Phase 4 (post-marketing). Regulators: FDA (US—BLA for biologics, NDA for small molecules); EMA (EU—CHMP opinion, marketing authorization); SFDA (Saudi Food and Drug Authority—pharmaceutical regulation in KSA, increasingly conducting independent review); MHRA (UK); PMDA (Japan). ICH GCP (Good Clinical Practice): global standard for conduct of trials—ethics, consent, data integrity, investigator responsibilities, sponsor responsibilities. Saudi site readiness: (1) SFDA CTA approval for any interventional trial; (2) IRB approval at each participating site; (3) qualified PI (board-certified, GCP-trained, trial experience); (4) study coordinator (research nurse typically); (5) regulatory and safety reporting infrastructure; (6) pharmacy for IP (investigational product) management; (7) source documentation (eSource increasingly); (8) monitoring visits; (9) audits. Saudi Phase 3 IBD trial participation has grown substantially—KFSHRC, KFMC, KAMC, KAUH, KSUMC, Dr Sulaiman Al Habib all enrolling in international Phase 2/3. Key challenges: slower regulatory start-up compared to Western sites; patient recruitment (smaller populations per site); translation and cultural adaptation of CRFs and informed consent; Ramadan fasting may affect visit timing; gender-matched research staff preferred. Opportunities: Saudi patients underrepresented in global IBD data—participation enriches evidence applicable to the region; SFDA increasingly recognized as rigorous; career advancement. Fellow role: shadow PI on ongoing trial, co-author site paper, eventually lead an investigator-initiated trial (IIT).
Saudi Arabia's participation in global IBD Phase 3 trials has doubled in the past 5 years. Fellows who build regulatory and site-readiness expertise now will become the site PIs of the next decade.
- CTA → Phase 1 → 2 → 3 → 4; ICH GCP global standard
- SFDA + IRB + GCP-trained PI = Saudi readiness
- Saudi IBD trial participation growing rapidly
- Challenges: start-up speed, recruitment, Ramadan, gender preferences
- Fellow path: shadow → site co-I → PI → IIT
- Modern IBD trial = induction + maintenance (treat-through) with composite endpoints
- NRI, central reading, ITT analysis = core methodology
- Saudi trial participation growing; SFDA increasingly rigorous
- Fellow path: shadow → site co-I → PI → investigator-initiated trial
QI Project Development Lab
Model for improvement, PDSA cycles, process mapping, aim statements — develop your own fellowship QI project.
1Model for Improvement & SMART Aim Statements
Model for Improvement (IHI): three questions + PDSA. (1) What are we trying to accomplish? (aim); (2) How will we know that a change is an improvement? (measures); (3) What changes can we make that will result in improvement? (change ideas). Then test with PDSA cycles. SMART aim: Specific, Measurable, Achievable, Relevant, Time-bound. Example poor aim: "Improve IBD care." Better aim: "Increase annual influenza vaccination rate among adults with IBD in our clinic from 45% to 80% by October 2027." Components: what, where, by when, by how much, from baseline. Three types of measures: (1) outcome measures (what ultimately matters—e.g., flare rate, hospitalizations, steroid use); (2) process measures (what we do—e.g., vaccination orders placed, calprotectin testing frequency); (3) balancing measures (unintended consequences—e.g., clinic visit length, patient satisfaction). Aim high but feasible—70–80% achievement rate keeps teams motivated. Saudi-specific examples: "Reduce TB screening gap before biologics from 35% to 5% in 6 months"; "Increase calprotectin testing in new Crohn's diagnoses from 40% to 90% in 1 year"; "Implement Arabic PHQ-9 at annual IBD review, achieve 90% completion in 6 months"; "Reduce pre-Hajj consultation miss rate from 70% to 20% by next Hajj season."
A QI aim without a specific number and a specific date is a wish, not an aim. Always include the from-to-by construction.
- Model for Improvement = 3 questions + PDSA
- SMART aim: Specific, Measurable, Achievable, Relevant, Time-bound
- Measures: outcome + process + balancing
- From-to-by-when structure mandatory
- Saudi examples: TB screening, calprotectin, Arabic PHQ-9, Hajj prep
2Process Mapping, Root Cause Analysis & Change Ideas
Process mapping: visualize current-state workflow with swim lanes (patient, nurse, physician, pharmacist, IT); identify steps, decision points, rework loops, waiting times, hand-offs, bottlenecks. Techniques: swim lane diagrams, value stream mapping, flowcharts. Root cause analysis (RCA): after identifying a problem, dig down to causes. Tools: (a) 5 Whys (keep asking why until you reach root cause); (b) Fishbone (Ishikawa) diagram with categories (people, process, equipment, materials, environment, measurement); (c) Pareto chart (80/20—identify the 20% of causes driving 80% of problems); (d) run charts (data over time showing trends and special-cause variation). Change ideas from root causes: (1) standardize (checklists, order sets, protocols—e.g., biologic initiation order set with auto-TB/HBV/pregnancy test); (2) automate (EMR alerts, auto-order, dashboards); (3) educate (patient and staff); (4) redesign (change flow—e.g., nurse-led vaccination review vs physician-dependent); (5) simplify (reduce steps); (6) redesign environment (visual cues). IHI change concepts (70 generic ideas)—use as inspiration. Key principle: small, iterative changes tested quickly beat grand plans. Saudi context: consider language (Arabic signage, scripts), staff turnover (easier-to-train standard work), EMR variation across centers (design for the lowest-resource environment), cultural fit (family involvement, religious accommodations).
Standard order sets in the EMR are the single highest-yield QI intervention for IBD. One well-built biologic initiation order set can eliminate most TB/HBV screening gaps overnight.
- Swim lane process mapping = current state visualization
- RCA tools: 5 Whys, fishbone, Pareto, run charts
- Change categories: standardize, automate, educate, redesign, simplify
- EMR order sets = high-yield intervention in IBD
- Saudi: Arabic scripts, staff training ease, EMR variation
3PDSA Cycles, Run Charts & Scaling Up
PDSA (Plan-Do-Study-Act): the fundamental cycle of improvement. Plan: specify the change, predict outcomes, define measures, who/what/when; Do: execute small test (e.g., 1 clinic day, 1 nurse, 5 patients), collect data, document surprises; Study: analyze against prediction, unexpected findings; Act: adopt, adapt, or abandon; plan next cycle. Small, rapid cycles (one-day tests) beat monthly committees. Sequential PDSA cycles build confidence and scale: cycle 1 = 1 patient → cycle 2 = 1 nurse day → cycle 3 = 1 clinic week → cycle 4 = all clinics → cycle 5 = spread to another site. Run charts: display data over time (x-axis = time, y-axis = metric); include median line; watch for shifts (6 consecutive points on one side of median), trends (5 points rising or falling), runs (too few or too many crossings), or astronomical points (obvious outliers); these indicate special-cause variation warranting investigation. Statistical Process Control (SPC) charts (e.g., p-chart, u-chart) add control limits and signal true improvement vs noise. Scale-up framework: "diffusion of innovations" (Rogers)—innovators (2.5%) → early adopters (13.5%) → early majority (34%) → late majority (34%) → laggards (16%); identify champions, provide visible wins, share data transparently, celebrate successes. Sustainability: embed in job descriptions, EMR workflows, training curricula, audits; avoid relying on individual enthusiasm. Saudi sustainability challenges: staff turnover (especially expatriate staff in some centers)—build knowledge into systems not people; secure leadership buy-in; integrate with CBAHI accreditation. SQUIRE 2.0 reporting: structure paper as Title/Abstract/Introduction/Methods (context, intervention, measures, analysis, ethics)/Results (run charts!)/Discussion (summary, interpretation, limitations)/Conclusion.
Your first PDSA should test with 1 patient or 1 clinic day, not a whole department rollout. Small tests reveal surprises early before they cause damage.
- PDSA = plan-do-study-act; small, rapid, iterative
- Run charts detect shifts, trends, runs, astronomical points
- SPC charts add statistical rigor
- Rogers diffusion: find champions, make wins visible
- SQUIRE 2.0 = reporting standard for QI publications
4Workshop: Develop Your Fellowship QI Project
Each fellow leaves this session with a one-page QI project charter. Template: (1) Problem statement (baseline data + why it matters); (2) SMART aim (from-to-by); (3) Scope (which patients, which clinic, which measure); (4) Team (PI, sponsor/program director, nursing lead, pharmacy, IT, patient representative when relevant); (5) Measures (outcome, process, balancing—with data source and collection plan); (6) Driver diagram (primary drivers → secondary drivers → change ideas); (7) PDSA cycles planned (at least 3); (8) Sustainability plan (what changes persist); (9) Timeline; (10) Ethics (IRB requirement varies—most QI is not research and doesn't need IRB, but documentation of QI classification advised; publishing may require IRB review—confirm with local policies). Examples of fellowship-level projects: (a) "Close the TB screening gap before biologic initiation" — baseline 35% → target 95% in 6 months via EMR order set + pharmacist verification; (b) "Increase calprotectin testing for treat-to-target" — from quarterly to 6-monthly per STRIDE-II; (c) "Implement Arabic PHQ-9 screening at annual review"; (d) "Reduce 30-day readmissions after ASUC discharge" — bundle of care including discharge checklist, 2-week follow-up call, early clinic visit; (e) "Standardize pre-Hajj consultation for IBD patients"; (f) "Improve transition readiness assessment using TRAQ in adolescents"; (g) "Cut door-to-biologic time in ASUC inpatients"; (h) "Increase SDM documentation for biologic selection." Submit charter to program director within 2 weeks; complete project within fellowship year; present at SGA annual meeting; submit for publication (Saudi J Gastroenterol, BMJ Open Quality).
The fellow who leaves this workshop without a written QI charter is the fellow who doesn't complete a QI project. Commit on paper before you leave the room.
- One-page QI charter due by end of workshop
- 10 elements: problem, aim, scope, team, measures, driver, PDSAs, sustain, timeline, ethics
- Examples span ASUC, screening, TB, Hajj, transition, PHQ-9
- Submit to PD within 2 weeks; complete by year-end
- Present at SGA meeting + submit for publication
- Model for Improvement + SMART aim + PDSA = structured QI
- Process mapping + RCA + standardization = intervention design
- Run charts detect real improvement vs noise
- Leave with signed QI charter — commit on paper
Journal Club: How to Read an IBD Trial
Structured critical appraisal of three recent pivotal IBD trials using CONSORT and risk of bias tools.
1CASP/CONSORT Framework for Critical Appraisal
Structured appraisal ensures no critical domain is missed. CASP (Critical Appraisal Skills Programme) RCT checklist: (1) Is the research question clear? (PICO); (2) Was assignment randomized? (sequence generation, allocation concealment); (3) Were groups similar at baseline? (Table 1 review); (4) Was blinding adequate? (participants, providers, outcome assessors); (5) Were all participants who entered the trial accounted for at the conclusion? (flow diagram, attrition); (6) Were participants, health workers, and study personnel blind? (7) Apart from the experimental intervention, were groups treated equally? (co-interventions); (8) How large was the treatment effect? (effect size, CI); (9) How precise was the estimate? (CI width); (10) Are the results applicable to your population? CONSORT 2010 (for reporting): 25-item checklist—title, abstract, introduction, methods (trial design, participants, interventions, outcomes, sample size, randomization, blinding, statistical methods), results (flow diagram, recruitment, baseline data, numbers analyzed, outcomes and estimation, ancillary analyses, harms), discussion (limitations, generalizability, interpretation), other information (registration, protocol, funding). Risk of bias (Cochrane RoB 2.0): five domains—randomization process, deviations from intended interventions, missing outcome data, measurement of outcome, selection of reported result; each rated low/some concerns/high. Quality of evidence (GRADE): starts high for RCT, downgrade for risk of bias, inconsistency, indirectness, imprecision, publication bias. Apply to every trial you read.
Journal clubs fail when they devolve into summaries. Force the group to rate each CONSORT item and each RoB domain — that's where learning happens.
- CASP 10 questions = structured appraisal
- CONSORT 25 items = reporting quality check
- RoB 2.0 = 5 domains for bias risk
- GRADE = evidence certainty downgrading
- Rate every item, don't just summarize
2Paper 1: Advanced Combination Therapy (VEGA-style)
Focus: combining two advanced therapies (e.g., guselkumab + golimumab in UC, or vedolizumab + adalimumab) to address the "therapeutic ceiling" observed with monotherapy (~40% endoscopic remission). Discussion questions: (1) What is the biological rationale (complementary MoA, e.g., anti-IL-23 + anti-TNF targets both innate/adaptive pathways)? (2) Who was eligible (severity, prior exposure)? (3) What were primary endpoints and how strict? (4) Were endpoints CONSORT-compliant? (5) How was safety captured—any excess infection or immunogenicity? (6) Subgroup interactions (biologic-naïve vs exposed)? (7) Effect sizes clinically meaningful? (8) Cost and access implications? (9) How does this change your practice today? (10) What Phase 3 follow-up is planned? Structured debate: two fellows present for combination therapy as future standard, two argue for monotherapy with treat-to-target as more cost-effective and safer; faculty moderates. Saudi context: cost is a major barrier—combination therapy prices would challenge formularies unless value demonstrated; SFDA approval for label combinations lags US/EU. Opportunity for registry analysis of off-label combinations in tertiary centers.
Combination therapy in IBD will eventually be standard for highest-risk patients — the question is how we identify them. Biomarker-guided selection is the research frontier.
- Combination therapy rationale: breaking 40% ceiling
- Complementary MoA (e.g., IL-23 + TNF)
- Safety signals require careful scrutiny
- Cost vs value frames Saudi applicability
- Biomarker-guided selection = research frontier
3Paper 2: Novel Oral Small Molecule (Upadacitinib / Etrasimod)
Focus: either upadacitinib (JAK-1 selective) in UC or CD, or etrasimod (S1P receptor modulator) in UC. Consider U-ACHIEVE/U-ACCOMPLISH or ELEVATE UC trials. Discussion questions: (1) MoA mechanism and selectivity rationale (JAK1 > pan-JAK to reduce VTE/infection risk); (2) Efficacy in bio-naïve vs bio-exposed; (3) Speed of onset—rapid remission is a marketing point but does it change practice? (4) Safety signals: VTE, MACE, herpes zoster, infections, hyperlipidemia; (5) Shingrix recommendation before tofacitinib/upadacitinib/etrasimod; (6) Sustained remission data at 52 weeks; (7) Cost vs biologics (oral may be cheaper or similar depending on setting); (8) Pregnancy/breastfeeding—most small molecules not recommended vs biologics; (9) Drug-drug interactions (CYP3A4 for upadacitinib)? (10) Role in treatment algorithm—second line after TNF or first line for certain phenotypes? Saudi practice question: how do you counsel a 28-year-old Saudi woman planning pregnancy about small molecule vs biologic? Answer: biologics preferred for pregnancy planning—strongly counsel against small molecules unless no alternative. Saudi context: etrasimod and upadacitinib both SFDA-approved; access and cost vary by insurance; patient preference for oral is significant.
Oral onset is a patient win but a pregnancy planning complication. Every reproductive-age woman on a JAK inhibitor needs contraception counseling and preconception plan.
- JAK selectivity (JAK1) reduces but doesn't eliminate VTE/MACE risk
- Shingrix before JAK inhibitors
- Not pregnancy-compatible — counsel reproductive-age women
- Oral convenience = adherence + preference
- Saudi SFDA approval but variable access
4Paper 3: Real-World/Saudi-Relevant Study
Focus: choose one of (a) Mosli 2023 Saudi IBD Registry results (phenotype, natural history, access); (b) a comparative effectiveness study from a Middle East cohort; (c) a pragmatic trial (e.g., PREVENT-CD, CALM, LIR!C); (d) a biosimilar switching study relevant to Saudi formulary; (e) a COVID-19 outcomes in IBD paper from Saudi/regional data. Discussion questions: (1) Study design (prospective cohort? retrospective? pragmatic trial?); (2) Representativeness of population—can it generalize to Saudi? (3) Confounding control (propensity scoring? regression? instrumental variables?); (4) Missing data and how handled; (5) Key findings—effect sizes, CIs; (6) Limitations acknowledged vs missed; (7) How does this fit with existing evidence (systematic review context)? (8) Regulatory/policy implications; (9) What gaps does this expose—future research ideas? (10) Would you design a Saudi study differently? This session should generate at least 3 research ideas fellows could pursue locally. Saudi paper example: Mosli et al. 2023 showed higher penetrating CD and earlier age of onset in Saudi cohort vs Western cohorts—implications for early aggressive therapy algorithms; also demonstrated delays in diagnosis and access gaps in non-tertiary centers. Use the analysis to frame a local registry QI project.
The Saudi IBD evidence base is thin — every journal club with a regional paper should end with at least one "we could answer this in Saudi" research idea.
- Real-world/Saudi studies anchor local generalizability
- Scrutinize confounding control and missing data
- Every journal club ends with ≥3 research ideas
- Mosli 2023 = phenotype, delay, access gaps — rich QI substrate
- Research ideas should drive fellowship QI/thesis projects
- CASP/CONSORT/RoB 2.0/GRADE = structured appraisal toolkit
- Combination therapy breaks the 40% ceiling but adds safety/cost considerations
- Oral small molecules convenient but not pregnancy-compatible
- Every journal club = ≥3 locally feasible research ideas
Assessment
QI project proposal presentation + Clinical trial critique + Research portfolio
Clinical Pearls
AGA Quality Measure #1: steroid-free remission — track this for every patient
PROs are increasingly required for trial endpoints — learn PRO-2, PRO-3, and IBDQ
OASIS and AMBER registries are SGA's contribution to real-world evidence
A well-designed QI project can be published — use SQUIRE 2.0 guidelines
Practice Points
Start a personal quality dashboard: steroid-free remission rates, screening compliance
Learn REDCap basics — standard tool for clinical registries used by OASIS and AMBER
Submit at least one abstract from your fellowship year
Key References
Kim AH, et al. ICHOM Standard Set for IBD. Lancet Gastroenterol Hepatol. 2022;7:115-126
Dulai PS, et al. AGA IBD Quality Measures 2.0. Gastroenterology. 2023;164:502-516
Mosli MH, et al. Saudi IBD Registry: Design and findings. Saudi J Gastroenterol. 2023;29:1-8
Reading List
ICHOM IBD Standard Set
Kim AH, et al. — Lancet Gastroenterol (2022)
AGA IBD Quality Measures 2.0
Dulai PS, et al. — Gastroenterology (2023)
How to critically appraise an IBD clinical trial
Colombel JF, et al. — J Crohns Colitis (2020)
REDCap for clinical registry management
Harris PA, et al. — J Biomed Inform (2009)