3.2 Adherence to medical advice
- Syllabus
- 9990–2028–2029
- Topic
- 3.2
- Level
- A2
Learning objectives
3.2.1Non-adherence types and reasons• 3.2.1 Types of non-adherence and reasons why patients do not adhere- types of non-adherence (failure to follow treatments and failure to attend appointments) and problems caused by non-adherence.- explanations of why patients do not adhere:- - rational non-adherence, including a study, e.g. Laba et al. (2012)- - Health Belief Model.- Relevant issues and debates and methodology for this topic include: application to everyday life, individual and situational explanations, reductionism versus holism, idiographic versus nomothetic, generalisations.3.2.2Measuring non-adherence• 3.2.2 Measuring non-adherence- subjective measures including clinical interviews and semi-structured interviews, including a study, e.g. Riekert and Drotar (1999).- objective measures focusing on pill counting and medication dispensers, including a study, e.g. Chung and Naya (2000).- biological measures including blood and urine samples.- Relevant issues and debates and methodology for this topic include: application to everyday life, idiographic versus nomothetic, quantitative and qualitative data, validity, reliability.3.2.3Improving adherence• 3.2.3 Improving adherence- improving adherence in children including a study, e.g. Chaney et al. (2004).- individual behavioural techniques: contracts, prompts, customising treatment.- community interventions (exemplified by the following key study).- Key study on improving medical adherence using community interventions: Yokley and Glenwick (1984).- Relevant issues and debates and methodology for this topic include: use of children in research, experiments, questionnaires, generalisations, validity.
Non-adherence can be intentional, unintentional and rational
| Behaviour | Examples | Problems |
|---|---|---|
| Treatment non-adherence | Never starts, wrong dose/time, stops early, changes diet/exercise, or uses device incorrectly | Reduced benefit, relapse/resistance, avoidable harm and misleading clinical decisions |
| Appointment non-attendance | Misses diagnosis, monitoring, treatment or follow-up | Delayed care, wasted capacity and disrupted continuity |
| Intentional vs unintentional | Deliberate preference/concern versus forgetting, misunderstanding, access or practical failure | The same observed missed dose needs a different intervention |
| Laba et al. (2012) | Evidence map |
|---|---|
| Aim/design | Online discrete-choice experiment testing which medication attributes influence stated continuation |
| Sample/task | 161 Australian community adults, age/gender matched to census; hypothetical medicines varied on eight attributes |
| Result | Immediate/long-term harms and benefits, cost and regimen significantly influenced choice; private insurance reduced cost sensitivity; side-effect framing often outweighed benefit framing |
| Boundary | Shows reasoned stated preference, not observed pill-taking; 10% participation and hypothetical non-disease-specific choices limit generalisation |
| Health Belief Model component | Adherence question / support |
|---|---|
| Susceptibility + severity | ‘What is my risk and what happens untreated?’ Give accurate, non-alarmist personalised information |
| Benefits | ‘How will this treatment reduce risk/symptoms?’ Make outcome and timescale concrete |
| Barriers/costs | Side effects, money, complexity, transport, stigma: solve or renegotiate them |
| Cues to action | Prompts, follow-up, symptom/action plans |
| Self-efficacy | Can the patient perform the regimen? Demonstrate, simplify and check skill |
First identify whether the barrier is preference, belief, memory, comprehension, access or regimen design. Shared decisions may customise timing/cost/side effects; information alone cannot fix transport or an unaffordable prescription.
‘Rational’ means understandable from the patient's perceived attributes and priorities, not medically optimal. HBM is useful but reductionist if family, service access, habit and practitioner communication are ignored.
Every adherence measure observes a proxy, so triangulate
| Measure | What it records | Strength | Validity threat |
|---|---|---|---|
| Clinical/semi-structured interview | Patient/caregiver account, routine and reasons | Rich context; semi-structure balances comparison and probing | Recall/social desirability; participating families may differ (Riekert & Drotar) |
| Pill count/refill | Tablets missing or prescriptions collected | Cheap, quantitative and practical | Dumping/sharing; collection does not prove ingestion |
| Electronic dispenser/TrackCap | Date/time bottle opened | Detailed objective event record | Opening ≠ swallowing; other containers and monitoring reactivity |
| Blood/urine | Drug/metabolite or expected biological marker | Harder to fake; confirms recent exposure for suitable medicines | Snapshot misses timing/occasional omission; metabolism varies; invasive and not universally available |
| Chung & Naya (2000) | Evidence map |
|---|---|
| Sample/regimen | 57 adults with asthma (32 male, 25 female; 18–55), 47 completed; zafirlukast 20 mg twice daily, 12 hours apart, away from meals |
| Procedure | After screening, 56 tablets supplied every three weeks across 12 weeks; bottle TrackCap recorded openings; returned pill count compared with cap data |
| Control/limit | Participants knew adherence was monitored but not exactly how. Standardised longitudinal records aid reliability, yet cap/pill proxies still do not prove ingestion and awareness can improve behaviour |
| Decision | Best response |
|---|---|
| Need reasons/barriers | Semi-structured interview |
| Need daily timing pattern | Electronic cap/dispenser |
| Need inexpensive clinic screen | Pill count/refill plus questions |
| Need recent physiological confirmation | Valid drug-specific blood/urine test with consent |
| High-stakes inference | Combine methods; investigate disagreement instead of declaring one ‘true’ |
Objective means less dependent on the researcher's or patient's judgement, not perfectly valid. A cap opening, missing pill or biomarker is evidence about a different part of adherence; agreement across methods strengthens inference.
Improve adherence by matching the intervention to the barrier
| Barrier | Technique/mechanism | Evaluation |
|---|---|---|
| Forgetting/weak cue | Timed text, call, sticker or specific prompt | Low cost; fatigue and ignored prompts reduce durability |
| Vague goal/accountability | Collaborative behavioural contract with exact action, monitoring, review and feasible reward | Supports commitment; honesty and clinician time constrain monitoring |
| Complex/unacceptable regimen | Customise timing, formulation, device, cost or side-effect plan through shared decision | Targets rational barriers; clinical safety/choice limits remain |
| Child finds treatment aversive | Developmentally appropriate device, play/reward and caregiver routine | Engagement may not equal correct dose; caregiver report is biased |
| Access/community barrier | Specific prompt plus transport/opening/access support or incentive | Can change attendance; cost, equity and external-reward maintenance matter |
| Chaney et al. (2004) Funhaler pilot | Evidence map |
|---|---|
| Intervention | Asthma spacer added toy-like incentive features to make correct use rewarding for young children |
| Method | Matched questionnaire-based survey; two home interviews of each caregiver by the same interviewer, plus random telephone checks about previous-day medication |
| Finding | Funhaler was associated with greater child/parent acceptance, willingness and reported adherence than the existing spacer |
| Limits | Small pilot, caregiver self-report and novelty; association/short follow-up cannot establish durable medication delivery or health improvement |
| Yokley & Glenwick (1984) | Evidence map |
|---|---|
| Sample/design | 1,133 families of under-immunised preschoolers randomly assigned by family to six conditions, avoiding different prompts within one household |
| Conditions | General prompt; specific prompt; specific+access; specific+monetary incentive; contact control; no-contact control |
| Result | All except general prompt improved immunisation; incentive largest, then access, then specific prompt. Specific prompt alone was most cost-effective |
| Evaluation | Random assignment and objective clinic behaviour support causality/application; one service context, incentives/equity and attendance versus broader adherence limit transfer |
Specify who receives the intervention, the exact cue/support/reward, the barrier it changes, an objective adherence outcome, baseline/comparison condition, follow-up and ethical safeguards. For children, obtain guardian consent and age-appropriate assent without coercive rewards or withholding necessary care.
The intervention with the largest immediate effect is not automatically best: compare cost, feasibility, equity, generalisation and maintenance. Fun or rewards can cue behaviour, but only valid measurement can show whether the prescribed regimen was actually followed.