Biostatistics & Population Health

Numeracy and health literacy in risk communication

Clinical Overview and When to Suspect Low Numeracy/Health Literacy

— Health literacy: capacity to obtain, process, and understand basic health information to make appropriate decisions (Healthy People 2030)

— Numeracy: ability to use and interpret quantitative information — probabilities, percentages, dosing, risk magnitudes

— Numeracy is a subdomain of health literacy but tracks independently; a patient may read fluently yet misinterpret "1 in 1000"

— ~36% of US adults have basic or below-basic health literacy (NAAL data)

— ~29% score at the lowest numeracy level; the elderly, Medicaid enrollees, those with <high-school education, and non-native English speakers are disproportionately affected

— Low numeracy is independent of IQ and overall literacy

— Forms returned incomplete or "I forgot my glasses"

— Patient identifies pills by color/shape rather than name

— Frequent missed appointments or medication errors

— Postpones decisions: "Whatever you think, doc"

— Poor adherence to warfarin, insulin, or any sliding-scale regimen — a numeracy red flag

— Inability to state when to take a "twice-daily" medication or convert "half a tablet"

— Risk communication failures drive nonadherence, preventable readmissions, malpractice claims, and disparities in screening uptake (mammography, colonoscopy, statin acceptance)

— Shared decision-making (SDM) legally and ethically requires the patient actually understand the numbers — not just hear them

Definitions

Epidemiology in the US

When to suspect limited literacy/numeracy in clinic

Why Step 3 cares

Board pearl: Universal precautions approach — assume every patient may have limited health literacy and tailor communication accordingly, rather than screening selectively. This is the AHRQ-endorsed standard and is the favored Step 3 answer when a question asks how to approach risk discussion with a "new patient" of unknown background.

Key distinction: Health literacy ≠ general literacy ≠ intelligence ≠ English proficiency — these are overlapping but separable constructs, and conflating them is a common distractor.

Presentation Patterns and Key History

— Discussing mammography starting at 40 vs 50: patient cannot weigh false-positive rates vs mortality benefit

— Statin for primary prevention: 10-year ASCVD risk of 9% — patient hears "low" and refuses

— Anticoagulation for AFib: CHA₂DS₂-VASc 3, HAS-BLED 2 — patient fixates on bleeding "chance"

— Prenatal aneuploidy screening: 1:250 quad-screen result misread as "I have Down syndrome"

— PSA screening: patient asks for the test because "my friend got cancer"

— Ask open-ended: "Tell me what you understand about your diabetes." Avoid "Do you understand?" — yes-bias is universal

— Medication reconciliation reveals doubled doses, skipped doses, or "I take it when I feel bad"

— Prior decisions made on anecdote rather than data ("My neighbor died on that drug")

— REALM (Rapid Estimate of Adult Literacy in Medicine) — reading recognition, ~2 min

— TOFHLA — comprehension, longer

— Newest Vital Sign (NVS) — uses an ice cream nutrition label, tests numeracy specifically, 3 min

— Single-Item Literacy Screener: "How confident are you filling out medical forms by yourself?" — answer of "a little/not at all" is sensitive

— "If a coin is flipped 1000 times, about how many heads?" (Schwartz numeracy item)

— "Which is the bigger risk: 1 in 100, 1 in 1000, or 1 in 10?"

— Inability is highly predictive of misunderstanding risk graphs

Clinical scenarios where numeracy gaps surface on Step 3

History clues

Validated screeners (know that they exist)

Numeracy-specific probes

Step 3 management: When the stem describes a patient who "nods along" but cannot restate the plan, the next best step is teach-back, not repeating the explanation louder or handing them a pamphlet. Teach-back is the single most testable closing technique in literacy questions.

Board pearl: Asking patients to bring all pill bottles ("brown bag review") is the quickest in-office surrogate for adherence and numeracy assessment.

Physical Exam Findings (and Communication Assessment)

— Patient hands forms to a family member to complete

— Reads slowly, traces text with a finger, skips sections

— Misuses medical terms ("sugar diabetes," "water pills") — not pathognomonic but a cue

— Avoids eye contact when written material is presented

— Excuses: "I left my glasses," "I'll read it at home," "My daughter handles that"

— Cannot demonstrate inhaler, glucometer, or insulin pen technique despite verbal claim of competence

— Teach-back ("show-me" method): "To make sure I explained this well, can you tell me in your own words how you'll take this medicine?"

— Demonstration: have patient draw up insulin, use peak-flow meter, or load an EpiPen trainer

— Read-back of dosing label: "Take 1 tablet twice daily" — ~46% of low-literacy adults cannot translate this into a correct timing plan

— Ask the patient to interpret their own A1c, BP log, or INR trend

— Show a pictograph (100 smiley faces, X colored to represent risk) — comprehension of icon arrays is far higher than percentages

— Test ability to convert "5% chance" → "5 out of 100"

— Note literacy concerns in the chart descriptively (e.g., "used teach-back; patient correctly restated plan after two attempts"), not as a label

— Avoid stigmatizing terms; literacy is a system failure, not a patient deficit

There is no physical exam for numeracy — but Step 3 will test the "communication exam": observable behaviors that substitute for a literacy screen

Observable signs during the encounter

Functional assessment

Numeracy-targeted bedside checks

Documentation

CCS pearl: On a CCS case where a patient is repeatedly nonadherent, ordering "patient education, medication" alone is insufficient — add teach-back verification and simplified written instructions at 5th–6th grade reading level, and schedule a short-interval follow-up (1–2 weeks) to confirm comprehension before escalating therapy.

Key distinction: Cognitive impairment (dementia, delirium) must be ruled out separately — MMSE/MoCA, not REALM, addresses cognition.

Diagnostic Workup — Framing Numbers Patients Can Actually Use

— Absolute risk (AR): "7 out of 100 people like you will have a heart attack in 10 years"

— Relative risk reduction (RRR): "Cuts your risk by 25%" — inflates perceived benefit

— Absolute risk reduction (ARR): "Lowers your risk from 8 to 6 out of 100" (2 percentage points)

— Number needed to treat (NNT): "50 people take the drug for 5 years for 1 to avoid an event"

— Number needed to harm (NNH): companion metric for adverse effects

— Always present absolute numbers, not relative

— Use consistent denominators ("out of 1000" for all comparisons — never mix 1/100 with 1/1000)

— Use frequencies, not percentages, when possible ("3 out of 100" beats "3%")

— Pictographs/icon arrays outperform bar charts and text for low-numeracy adults

— Provide both benefits and harms side-by-side ("balanced framing")

— Avoid loss-framing alone ("you'll die without this") — pair with gain frame

— Denominator neglect: "1 in 10" feels bigger than "10 in 100"

— Framing effect: "90% survival" accepted more than "10% mortality" — same data

— Ratio bias: larger numerators feel scarier (24/1000 > 4/100 to many patients)

— Availability heuristic: a vivid anecdote outweighs base rates

Core formats for communicating risk

Evidence-based principles (IOM, Cochrane, Fagerlin frameworks)

Common cognitive pitfalls to anticipate

Board pearl: When a Step 3 stem asks the best way to convey a screening test's benefit, the answer is almost always absolute risk reduction expressed as a natural frequency with a consistent denominator, ideally with an icon array — not RRR, not "p < 0.05," not NNT alone.

Key distinction: RRR is the marketer's number; ARR/NNT is the clinician's number; natural frequencies are the patient's number.

Diagnostic Workup — Interpreting Screening Test Probabilities

— Classic example: A mammogram has 90% sensitivity, 91% specificity. Prevalence of breast cancer in a 40-year-old = ~0.8%. If positive, what is the probability she has cancer?

— Most physicians say 80–90%. Correct answer: ~7–9% (PPV)

— Reason: low pretest probability dominates; most positives are false positives

— Patients (and exam stems) routinely confuse sensitivity with PPV

— A positive result is not "a 90% chance you have the disease"

— Counseling must reframe: "Out of 100 women your age with a positive mammogram, about 9 will actually have cancer; the other 91 will not"

— Use natural frequency trees — start with 1000 people, branch by disease status, then by test result

— Example for prenatal screening: 1000 pregnancies → 1 has trisomy 21 → quad screen positive in that 1 plus ~50 of the unaffected → PPV ~2%

— This single visual converts "1 in 250 result" anxiety into appropriate next-step thinking (offer cfDNA or diagnostic amnio)

— Avoid "p-values" entirely; translate CIs as "the true benefit is somewhere between X and Y, most likely around Z"

— Wide CI → "we are not sure yet"; narrow CI → "we are confident"

Bayesian reasoning patients (and physicians) get wrong

Why this matters for Step 3

Communicating uncertainty

Confidence intervals for the patient

Step 3 management: When a screening result comes back positive and the patient is panicking, the next best step is to explain the PPV at this prevalence using a natural frequency before ordering confirmatory testing. This both calms the patient and supports informed consent for the next (often invasive) step.

Board pearl: Gigerenzer's "natural frequencies" approach raises correct Bayesian reasoning in physicians from ~10% to ~50% — and is the format Step 3 rewards.

Key distinction: Sensitivity/specificity are test properties; PPV/NPV are patient-relevant and depend on prevalence.

Risk Stratification — Matching Communication to Patient Need

— Tier 1 — Universal precautions: plain language, teach-back, limited concepts per visit (≤3), absolute numbers, written backup at 5th–6th grade level — applied to every patient

— Tier 2 — Targeted: for patients flagged by screener, observation, or high-stakes decision (anticoagulation, oncology, end-of-life) — add icon arrays, decision aids, longer visit, family involvement

— Tier 3 — Intensive: for major irreversible decisions (surgery, transplant, hospice) — formal shared decision-making with validated decision aid, multiple visits, interpreter if needed

— PSA screening (USPSTF Grade C, 55–69) — SDM is required

— Lung cancer screening LDCT (50–80, 20 pack-yr, current or quit <15 yr) — SDM is a CMS reimbursement requirement

— AFib anticoagulation

— Mastectomy vs lumpectomy + radiation

— Elective PCI for stable angina (per COURAGE/ISCHEMIA era counseling)

— Increase knowledge, improve risk perception accuracy, decrease decisional conflict

— Do not increase anxiety; often decrease elective procedure uptake when patients learn true marginal benefit

— Avoid percentages with denominators >100

— Always pair number with comparator ("about the same as your risk of…")

— Anchor to familiar referents (size of a stadium, deck of cards)

Tiered approach to risk communication

High-stakes scenarios on Step 3 favoring decision aids

Decision-aid evidence (Cochrane)

Numeracy-tailored elements

Board pearl: When a stem asks the most appropriate next step before ordering PSA or LDCT, the answer is shared decision-making with a decision aid, not "order the test" or "decline the test."

CCS pearl: Documenting "SDM performed, decision aid reviewed, patient elects [option]" is both clinically and medicolegally protective and is favored CCS phrasing for elective screening.

Interventions — First-Line Communication Techniques

— Replace jargon: "hypertension" → "high blood pressure"; "lipids" → "blood fats"; "benign" → "not cancer"

— Short sentences, active voice

— Chunk and check: 2–3 key points per encounter, verify each with teach-back

— Avoid "negative" framing ("don't forget") — use positive ("take with breakfast")

— Provider says: "I want to make sure I did a good job explaining. Can you tell me how you'll take this medicine when you get home?"

— Patient restates → provider corrects gaps → patient restates again until accurate

— Studies: improves adherence, reduces 30-day readmissions in HF, improves A1c

— Frames error as the provider's, not the patient's — preserves dignity

— Ask what the patient already knows/wants to know

— Tell in plain language, ≤3 chunks

— Ask for restatement (teach-back) — closes the loop

— Target 5th–6th grade reading level (AMA/NIH recommendation)

— Tools: Flesch-Kincaid, SMOG readability formulas

— Large font (≥12 pt), high contrast, white space, illustrations

— Bullet points beat paragraphs

— Icon arrays/pictographs for risk (100 figures with proportional shading)

— Bar graphs for comparing two options

— Avoid pie charts for small risks (<10%) — wedges are unreadable

— Required by Title VI of the Civil Rights Act and Section 1557 of the ACA for any federally funded provider

— Use certified medical interpreters — not family members, not minors, not bilingual staff acting ad hoc

— Telephonic/video interpretation acceptable when in-person unavailable

Plain language principles

Teach-back method (the most testable single technique)

Ask-Tell-Ask

Written materials

Visual aids

Interpreter services

Step 3 management: A patient with limited English proficiency requests their adult son interpret for an oncology consent. The correct next step is to decline and arrange a certified medical interpreter, even if the patient consents to family use, except in emergencies.

Board pearl: Teach-back works in <2 minutes and is the single highest-yield literacy intervention tested.

Advanced Tools — Decision Aids, Risk Calculators, and Systems-Level Interventions

— Structured tools that present options, outcomes, probabilities, and elicit values

— Must meet IPDAS (International Patient Decision Aid Standards) criteria

— Examples: Mayo Statin Choice, Option Grids, Ottawa Personal Decision Guide, AHRQ SHARE tools

— Best deployed before or during the visit, not handed at checkout

— ASCVD Pooled Cohort Equations — outputs 10-year and lifetime risk; patients struggle with "10-year" timeframe — translate to "if 100 people just like you…"

— CHA₂DS₂-VASc / HAS-BLED — present both stroke and bleed risk side by side; not stroke risk alone

— FRAX for osteoporosis, Gail for breast cancer, MELD for liver — same principle: convert to natural frequencies before presenting

— Health literacy universal precautions toolkit (AHRQ) — implementation guide for practices

— Pictogram-based medication labels (USP standard, e.g., "Take 1 tablet by mouth in the morning and 1 tablet by mouth at bedtime")

— Pill organizers, blister packs, once-daily formulations — reduce numeracy load

— Medication reconciliation at every transition — TJC National Patient Safety Goal

— Discharge summaries written at 6th-grade level with teach-back at discharge — reduces 30-day readmission

— Patient portals: often written at 10th–12th grade — a literacy barrier, not a solution

— Apps and SMS reminders improve adherence in low-numeracy populations when paired with simple visuals

— Beware digital divide — equity issue, especially elderly and rural

Patient decision aids (PtDAs)

Risk calculators and their communication challenges

Systems-level interventions

Technology

CCS pearl: On a CCS discharge for a patient with new warfarin or insulin, the high-yield orders are: patient education with teach-back, simplified written instructions, pillbox or insulin pen with demonstration, anticoagulation/diabetes nurse referral, and 1-week follow-up — bundling these is what scores.

Board pearl: Decision aids reduce elective procedure rates and improve knowledge — the favored exam answer for "what does adding a decision aid do?"

Special Populations — Elderly and Cognitive/Sensory Decline

— Adults ≥65 have the lowest health literacy scores of any age group (NAAL)

— Compounded by presbyopia, presbycusis, mild cognitive impairment, polypharmacy

— ~40% of community-dwelling elders make at least one medication error monthly

— Smaller font on Rx labels and AVS (After Visit Summary)

— Multiple prescribers → conflicting instructions

— Complex regimens (>5 meds, multiple daily doses)

— Difficulty interpreting "as needed," "with food," sliding-scale insulin

— Larger print (≥14 pt), high-contrast materials

— Quiet room, face the patient, lower-pitched voice (presbycusis affects high frequencies first)

— One topic per visit when feasible; written summary to take home

— Involve caregiver with patient's consent (HIPAA-compliant) — but address the patient first

— Deprescribe to reduce numeracy burden (Beers Criteria, STOPP/START)

— Pill organizers, blister packs, once-daily dosing preferred

— Home health nurse visit for new insulin, anticoagulant, or inhaler

— If teach-back fails repeatedly despite plain language, screen with Mini-Cog or MoCA before assuming literacy alone

— Distinguish literacy (knowledge gap, modifiable with format) from MCI/dementia (executive dysfunction, requires caregiver-centered planning)

— Capacity is decision-specific and may fluctuate

— Low literacy alone does not equal lack of capacity

— A patient can have low numeracy yet retain full capacity if they can express choice, understand, appreciate, and reason about their situation

Epidemiology

Specific barriers

Modifications

Cognitive screening

Decision-making capacity

Key distinction: Capacity assessment evaluates the process of decision-making, not the "correctness" of the decision. A patient declining a statin after a competent SDM discussion is exercising autonomy, not lacking capacity.

Step 3 management: An 82-year-old with low numeracy and 8 medications is the prototypical Step 3 candidate for comprehensive medication review with deprescribing and caregiver-inclusive teach-back at the next visit.

Special Populations — Pediatrics, Pregnancy, LEP, and Disability

— Pediatric medication errors disproportionately affect parents with low numeracy (misreading mL vs tsp, weight-based dosing)

— Always dispense liquid meds with a dosing syringe, never a kitchen spoon

— Use mL only (the 2015 AAP policy) — eliminates teaspoon/tablespoon confusion

— Demonstrate the dose, then have parent demonstrate back

— Tools: HELP project materials, Health Literacy Universal Precautions for pediatrics

— Confidentiality discussions about contraception, STIs, mental health require developmentally appropriate plain language

— Assess teen's own literacy separate from parent's

— Prenatal screening (cfDNA, quad screen, NT) is a numeracy minefield

— Always present results as natural frequencies with prior risk × likelihood ratio framing

— Decision aids for VBAC, prenatal genetic testing, GBS prophylaxis improve informed consent

— Certified medical interpreter required (Title VI, ACA §1557)

— Family/friends, minors, ad-hoc bilingual staff are inappropriate except in true emergencies — and even then must be documented

— Written materials in patient's language, not English handed through interpreter

— Cultural concordance ≠ language concordance — both matter

— Visual impairment: large print, audio summaries, screen-reader-compatible portals

— Hearing impairment: ASL interpreter (in-person or VRI), not lip-reading or written notes alone for complex discussions — required under ADA

— Cognitive/intellectual disability: plain language, supported decision-making, guardian involvement per state law

Pediatrics — parental literacy is the lever

Adolescents

Pregnancy

Limited English proficiency (LEP)

Disability

Board pearl: For LEP patients, using a family member as interpreter is the classic wrong-answer trap — even if patient and family both "prefer" it, the standard of care and legal requirement is a certified interpreter for clinically significant encounters.

Key distinction: Health literacy demands are highest at transitions — birth, new diagnosis, hospital discharge, end of life — these are the Step 3 hot zones.

Complications and Adverse Outcomes of Poor Risk Communication

— Medication errors: wrong dose, wrong frequency, dangerous duplications — leading cause of preventable outpatient adverse events

— Nonadherence: 50% of chronic-disease medications are not taken as prescribed; literacy/numeracy is a major modifiable driver

— Missed screening: lower mammography, colonoscopy, and vaccination rates in low-literacy populations

— Overuse: patients with low numeracy more likely to accept unnecessary screening (PSA, whole-body CT) due to misperceived benefit

— Underuse: declining statins, anticoagulants, vaccines based on misunderstood risk

— Worsened outcomes: higher A1c, BP, HF readmissions, all-cause mortality in low-literacy cohorts (independent of SES)

— Warfarin/DOACs: dose changes, INR interpretation, bleed/clot recognition

— Insulin: unit confusion, sliding scale errors, hypoglycemia recognition

— Methotrexate: weekly vs daily dosing — has caused fatal overdoses

— Opioids: "as needed" misinterpretation, overdose risk

— Pediatric liquids: 10-fold dosing errors

— Higher 30-day readmission rates — CMS HRRP penalties

— Increased ED utilization for ambulatory-sensitive conditions

— Disparities in chronic disease control

— Malpractice exposure — failure of informed consent

— Low health literacy is concentrated in marginalized populations and amplifies existing health disparities

— Addressing literacy is a structural intervention, not a patient-blaming exercise

Patient-level harms

Specific high-risk medication scenarios

Systems-level harms

Equity dimension

Board pearl: The strongest independent predictor of medication error among older adults in the outpatient setting is low health literacy/numeracy, exceeding number of medications and number of prescribers in many studies.

Step 3 management: When a stem describes recurrent hypoglycemia in an insulin-using elderly patient with low literacy, the highest-yield next step is simplifying the regimen (basal-only or fixed-dose), demonstrating with teach-back, and referring to a diabetes educator — not titrating to a tighter A1c.

When to Escalate — Consults, Programs, and System Resources

— ≥5 chronic medications, multiple prescribers, recent transition of care

— New anticoagulant, insulin, immunosuppressant, or chemotherapy

— Repeated dosing errors or adherence failures

— Pharmacist-led medication therapy management (MTM) is reimbursed under Medicare Part D for eligible beneficiaries

— New T1DM/T2DM diagnosis

— Insulin initiation

— A1c not at goal despite intensified regimen

— Hypoglycemia unawareness

— Medicare covers DSMT (Diabetes Self-Management Training) — 10 hours initial, 2 hours/year ongoing

— Housing, food insecurity, transportation barriers compound literacy issues

— Insurance/coverage gaps blocking medication access

— Need for home health, durable medical equipment, or community resources

— Anxiety/depression interfering with comprehension or adherence

— Health literacy + low motivation suggests motivational interviewing is indicated

— Recurrent admissions for the same ambulatory-sensitive condition (HF, COPD, DKA) + identified literacy barrier → transitional care intervention (e.g., Coleman Care Transitions, Project BOOST, Project RED) at discharge

— These bundles include: simplified med list, teach-back, follow-up call at 48–72 hours, PCP appointment within 7 days

— New insulin, new anticoagulant, complex wound care, oxygen, tube feeding, recent stroke — particularly when literacy is limited

— Reduces 30-day readmission

Refer to a clinical pharmacist when

Refer to a certified diabetes care and education specialist (CDCES) when

Refer to social work / care coordination when

Refer to behavioral health when

Inpatient triggers

Home health nurse referral

CCS pearl: On a discharge case with a low-literacy patient on new high-risk medications, the highest-yield orders to add are pharmacist medication reconciliation, home health nurse, follow-up phone call in 48–72 hours, and PCP appointment within 7 days — this bundle is the evidence-based readmission-reduction package.

Board pearl: Recurrent HF admissions in a patient with low numeracy who cannot interpret daily weights → HF disease management program with home health and teach-back is the favored Step 3 answer over uptitrating diuretic.

Key Differentials — Same-Category (Communication-Related) Causes of Nonadherence

— Patient's framework for illness differs from biomedical model

— Use Kleinman's questions: "What do you think caused it? What does it do to you? What do you fear most?"

— Intervention: cultural humility, negotiation of treatment plan

— Historical and ongoing medical mistrust (Tuskegee, Henrietta Lacks, ongoing disparities) — particularly relevant in Black, Indigenous, and immigrant communities

— Intervention: build trust over time, acknowledge legitimacy of concerns, do not dismiss

— Patient understands risk but is not ready to change (precontemplation per Stages of Change)

— Intervention: motivational interviewing, not more education

When a patient appears nonadherent or misunderstands risk, don't default to literacy — work through the differential of communication failures

Low health literacy — knowledge/skill gap; responsive to plain language, teach-back, decision aids

Low numeracy specifically — may have normal reading; struggles with probabilities, dosing math; responsive to natural frequencies, icon arrays, simplified regimens

Language barrier (LEP) — primary issue is translation, not comprehension capacity; intervention is certified interpreter, not simpler English

Sensory impairment — vision (presbyopia, macular degeneration, diabetic retinopathy) or hearing (presbycusis); intervention is sensory accommodation, not literacy materials

Cultural/explanatory model mismatch

Health beliefs and trust

Motivational/behavioral

Cognitive impairment — see chunk 9; assess separately

Mental health — depression reduces adherence ~3-fold; anxiety distorts risk perception

Board pearl: "Noncompliant" is a discouraged term — the contemporary framing is "nonadherent," and the differential above must be worked through before labeling. Step 3 will reward the answer that explores the reason ("ask the patient why") over the answer that escalates therapy.

Key distinction: Knowledge deficit (literacy) needs education; motivation deficit needs motivational interviewing; trust deficit needs relationship; structural barrier (cost, transportation) needs social work — pick the right tool.

Key Differentials — Structural and System-Level Causes

— ~25% of US adults report skipping doses, splitting pills, or not filling Rx due to cost

— Especially with insulin, inhalers, DOACs, oncology drugs

— Intervention: generic substitution, $4 lists, patient assistance programs, 340B, Medicare Extra Help (LIS), pharmacy benefits manager review

— Rural, elderly, disabled patients may not reach the pharmacy or clinic

— Mail-order pharmacy, telehealth, community health workers

— Medicaid non-emergency medical transportation benefit

— Coverage changes mid-treatment cause formulary disruption, prior auth delays, gaps in care

— Particularly disruptive for transitions (hospital discharge, post-partum Medicaid loss)

— Multiple specialists, multiple EMRs, conflicting recommendations

— Intervention: medical home model, care coordinator, shared care plan

— Rushed visits (<15 min), failure to use teach-back, jargon, no written summary

— Mismatched expectations about who follows up on results

— Test result communication failures are a top malpractice claim

— Most adverse drug events occur within 30 days of discharge

— Causes: incomplete med rec, no follow-up appointment, AVS at 12th-grade reading level, no teach-back

— Solution bundles: Project RED, Coleman, BOOST (see chunk 12)

— Patient portal use lowest among elderly, low-income, rural, non-English speakers — the same populations with lowest health literacy

— Do not assume MyChart messaging is adequate follow-up

Cost-related nonadherence

Access and transportation

Insurance churn

Health system fragmentation

Provider-side communication failures

Discharge transition failures

Digital divide

Step 3 management: When the stem reveals a patient skipping insulin doses to "make it last to the end of the month," the next best step is screening for cost barriers and connecting to patient assistance / formulary alternatives — not patient education on adherence.

Board pearl: Always ask "How are you paying for your medications?" — cost-related nonadherence is invisible unless asked about explicitly.

Long-Term Plan — Embedding Literacy into Chronic Care

— Plain language and teach-back at every visit, not just new diagnoses

— Reassess understanding at each medication change, each abnormal result, each transition

— Document literacy-related accommodations in the chart

— Diabetes: annual DSME refresher, glucometer/CGM technique check, insulin pen demo, sick-day rules in plain language, hypoglycemia action plan with pictures

— HF: daily weight log with action thresholds ("if up 3 lb in 2 days, call us"), low-sodium diet handout at 5th-grade level, medication "zone" tool (green/yellow/red)

— Asthma/COPD: written action plan with color zones; inhaler technique reassessed at every visit (most patients use inhalers incorrectly even years in)

— Anticoagulation: anticoag clinic follow-up, INR log if warfarin, bleed/clot warning signs with examples, drug-drug/diet interaction handout

— Hypertension: home BP log with target range, action plan for high readings

— Annual comprehensive med review

— Deprescribing per Beers/STOPP at each transition

— Pill organizer or blister pack

— Synchronize refills to one date ("med sync")

— With patient consent, identify a "health partner" who attends visits and shares plan

— Especially important in dementia, complex chemo, transplant

— Motivational interviewing for sustained change

— Group visits and peer support (e.g., diabetes group medical visits)

— Patient navigator programs for cancer, transplant, complex chronic disease

— Community health workers / promotoras — proven to improve outcomes in low-literacy populations

Universal precautions as standing practice

Chronic disease–specific plans

Polypharmacy management

Family/caregiver integration

Behavioral support

Health system supports

Board pearl: The action plan with color-coded zones (green = stable, yellow = caution, red = call/ED) is the prototype low-literacy chronic disease tool — high-yield for asthma, HF, and anticoagulation on Step 3.

Step 3 management: At each chronic care visit, reassess understanding of warning signs and action steps with teach-back — not just labs and medications.

Follow-Up, Monitoring, and Counseling Cadence

— New high-risk medication (insulin, warfarin, DOAC, opioids, methotrexate): phone or in-person follow-up within 1–2 weeks

— Post-hospital discharge: call within 48–72 hours, clinic visit within 7 days (sooner for HF, COPD exacerbation)

— New chronic diagnosis: follow-up within 2–4 weeks for second teaching pass

— Stable chronic disease: quarterly to annual per condition; always reassess literacy at transitions

— Adherence: refill records, pill counts, MPR (medication possession ratio), self-report with non-judgmental framing ("Most people miss doses sometimes — how often does that happen for you?")

— Comprehension: teach-back at each visit on the highest-stakes element (warning signs, dose changes)

— Outcomes: disease-specific markers (A1c, BP, INR, daily weight, peak flow)

— Adverse events: specifically asked, not waited for

— Use motivational interviewing for behavior change: open-ended questions, affirmations, reflective listening, summaries (OARS)

— Use 5 A's for behaviors (smoking, alcohol, weight): Ask, Advise, Assess, Assist, Arrange

— Avoid lecturing; elicit the patient's reasons for change

— Document literacy-related interventions: "Teach-back used; patient correctly restated insulin dosing on second attempt"

— Document SDM: "Reviewed [decision aid]; patient understands ~7% 10-year ASCVD risk and elects statin / declines statin"

— Document interpreter use: name/ID, language, modality

— CAHPS surveys include communication items

— HCAHPS "communication about medications" is a publicly reported hospital metric

— Value-based care contracts increasingly include health literacy–sensitive measures

Visit cadence by risk

What to monitor specifically

Counseling principles

Documentation standards

Quality metrics

Board pearl: A 48–72-hour post-discharge phone call is one of the highest-yield, lowest-cost interventions to reduce 30-day readmissions — a frequent Step 3 correct answer for the right scenario.

CCS pearl: When closing a CCS case, always include follow-up appointment at appropriate interval and patient education/counseling orders — omission costs points even if pharmacotherapy is perfect.

Ethical, Legal, and Patient Safety Considerations

— Three pillars: disclosure, capacity, voluntariness — disclosure fails if the patient cannot understand the numbers

— A signed consent form is not legally protective if the patient never understood the risks — courts apply the reasonable patient standard (what a reasonable patient would want to know)

— Document teach-back as evidence of comprehension, especially for high-risk procedures

— Edge case: A patient with low numeracy signs consent for elective surgery but cannot restate the major risks. The ethically correct next step is to re-explain using plain language and visuals, reassess, and only proceed if comprehension is demonstrated — even if it delays the case.

— Required (CMS) for lung cancer screening LDCT

— Ethically required for any preference-sensitive decision: PSA, mammography 40–49, statin primary prevention, AFib anticoagulation, mastectomy vs lumpectomy

— SDM is not "patient chooses alone" — it is shared, with the clinician offering expertise and the patient offering values

— Plain-language reporting of medical errors is required (TJC, state law) — patients must be told what happened, what it means, what's next in language they understand

— Discharge AVS at high reading level + no teach-back + new high-risk meds = predictable harm; this is a patient safety event by NQF and TJC standards

— Failure to ensure understanding at discharge has been the basis for malpractice judgments

— Section 1557 of ACA prohibits discrimination by language, disability, race

— Failure to provide interpreters or accessible materials = civil rights violation, not merely poor service

— Family interpreters may compromise confidentiality (reproductive, mental health, IPV)

— Always offer the patient a private discussion with a professional interpreter before involving family

Informed consent requires actual understanding

Shared decision-making and preference-sensitive care

Mandatory reporting and disclosure

Transition-of-care risk

Equity and nondiscrimination

Confidentiality vs family involvement

Board pearl: Consent obtained through an ad-hoc family interpreter for a non-emergent procedure is not valid informed consent and exposes the clinician and institution to liability — the correct answer is to reschedule with a certified interpreter.

Key distinction: Autonomy ≠ abandonment — clinicians must ensure patients have the information and comprehension needed to exercise autonomy meaningfully.

High-Yield Associations and Rapid-Fire Clinical Facts

~36% of US adults have basic or below-basic health literacy (NAAL)

~29% score in the lowest numeracy level

Health literacy is independent of overall literacy and IQ

Teach-back is the single highest-yield bedside technique — frames error as provider's, not patient's

5th–6th grade is the target reading level for patient materials (AMA/NIH)

Plain Writing Act of 2010 requires federal agencies to use plain language

Absolute risk > relative risk for patient communication

Natural frequencies (3 out of 100) > percentages (3%) > probabilities (0.03) for comprehension

Consistent denominators are essential — never mix "1 in 100" with "10 in 1000"

Icon arrays/pictographs outperform bar charts and text for risk perception

Avoid pie charts for small risks

Loss framing ("10% will die") feels worse than gain framing ("90% survive") for identical data — framing effect

Denominator neglect: patients often judge 1/10 < 10/100

RRR inflates perceived benefit — drug marketing's favorite metric

NNT is intuitive once explained; NNH is its safety counterpart

Sensitivity ≠ PPV — PPV depends on prevalence; classic mammography example

Bayesian frequency tree is the format that fixes physician-level Bayes errors

Certified medical interpreters required for LEP — Title VI, ACA §1557 — family/minors are inappropriate

mL only for pediatric liquid dosing (2015 AAP)

Methotrexate weekly — fatal if taken daily; literacy-sensitive labeling required

Color-zone action plans for asthma, HF, anticoagulation

48–72-hour post-discharge call reduces readmissions

Medicare DSMT: 10 hr initial, 2 hr/year for diabetes self-management

CMS requires SDM for lung cancer screening LDCT

USPSTF Grade C for PSA (55–69) — requires SDM

HCAHPS publicly reports hospital communication metrics

Brown-bag review = bring all pill bottles; fastest in-office numeracy/adherence screen

Newest Vital Sign (NVS) = numeracy-focused screener using an ice cream label

Board pearl: If two answer choices both seem reasonable, the one that involves the patient (SDM, teach-back, ask the patient why) typically beats the one that acts unilaterally.

Key distinction: Numeracy is to risk what literacy is to text — both must be addressed for true informed consent.

Board Question Stem Patterns

— A 55-year-old asks about her statin benefit. ASCVD 10-yr risk is 12%; statin reduces RR by 25%.

— Right answer: "In 10 years, about 12 of 100 people like you would have a heart attack or stroke without the medicine, and about 9 of 100 with the medicine."

— Wrong answers: "It cuts your risk by 25%" (RRR, inflates); "Your p-value is significant" (meaningless to patient); "Just trust me" (paternalism)

— Diabetic patient on new insulin nods through teaching but later presents with hypoglycemia from doubled doses.

— Right answer: Use teach-back at the next visit and simplify the regimen.

— Wrong answers: Hand a more detailed pamphlet; switch to a different insulin without addressing comprehension.

— LEP patient brings adult son to interpret consent for elective surgery.

— Right answer: Use a certified medical interpreter (in-person, telephonic, or VRI).

— Wrong answers: Proceed with son interpreting; postpone indefinitely; document patient declined interpretation.

— Patient receives a positive mammogram and asks if she has cancer.

— Right answer: Explain PPV using natural frequencies ("of 100 women with a positive result, about 9 actually have cancer"), then arrange diagnostic follow-up.

— Wrong answers: Reassure without numbers; quote sensitivity; order diagnostic test without explanation.

— 60-year-old man asks about PSA.

— Right answer: Shared decision-making with a decision aid.

— Wrong answers: Order PSA; refuse to discuss; order without SDM.

— Elderly HF patient with low literacy readmitted thrice.

— Right answer: Transitional care bundle — teach-back at discharge, simplified action plan, home health, 48–72 hr call, 7-day follow-up.

— Wrong answers: Increase diuretic dose alone; refer to cardiology only.

— Patient splitting insulin doses to save money.

— Right answer: Screen for cost, connect with assistance, switch to affordable alternative.

Stem pattern 1 — "Best way to explain risk"

Stem pattern 2 — "Patient nods but doesn't understand"

Stem pattern 3 — "Family member as interpreter"

Stem pattern 4 — "Positive screening test panic"

Stem pattern 5 — "Preference-sensitive screening"

Stem pattern 6 — "Recurrent readmissions"

Stem pattern 7 — "Cost-related nonadherence"

Board pearl: When the stem mentions a patient's behavior (nonadherence, anxiety, declining treatment), the right answer almost always starts with the patient — ask, teach-back, SDM — before changing therapy.

One-Line Recap

— Format matters: absolute risk > relative risk; natural frequencies (3 of 100) > percentages > probabilities; consistent denominators; icon arrays beat words; avoid pie charts for small risks; pair gain and loss framing.

— Technique matters: universal precautions for every patient; plain language at 5th–6th grade level; ≤3 key points per visit; teach-back to verify; Ask-Tell-Ask for new information; motivational interviewing for behavior change; certified medical interpreters (never family/minors) for LEP.

— Decisions matter: preference-sensitive choices (PSA, LDCT lung screening, AFib anticoagulation, statin primary prevention, mastectomy vs lumpectomy) require shared decision-making with a validated decision aid; document the conversation; PPV (not sensitivity) is what patients need after a positive screen, explained via a Bayesian frequency tree.

— Systems matter: transitions of care are the highest-risk literacy moments; bundle simplified discharge instructions, teach-back, home health when indicated, 48–72-hour follow-up call, and clinic visit within 7 days; screen for cost-related nonadherence; deprescribe to reduce numeracy load in the elderly; color-zone action plans for asthma, HF, and anticoagulation.

Core teaching point: Effective risk communication in Step 3 means assuming universal limited literacy and numeracy, presenting probabilities as absolute natural frequencies with consistent denominators (often supported by icon arrays and decision aids), and closing the loop with teach-back so that informed consent and shared decision-making are real, documentable, and equitable.

High-yield recap bullets

Board pearl: When two answers are clinically equivalent, choose the one that engages the patient (teach-back, SDM, ask the patient why) over the one that acts on the patient — this is the through-line of every numeracy and health literacy question on Step 3.

Key distinction: Health literacy is a system property, not a patient deficit — fix the communication, not the patient.