The action plan that didn't land
A diabetes clinic serves 200 adults with HbA1c above 9%. After one 15-minute visit, each patient receives a detailed 12-week plan: a specific diet, an exercise schedule, a pill-tracking app, and a glucose log. At 8 weeks, 31 patients are following the plan; 140 have not opened the app; 29 did not return. Follow-up interviews find that most non-adherent patients say their doctor 'doesn't get it' and describe the plan as 'one more thing.' Several report that after missing two days of logging in week two, they stopped entirely, saying 'I already blew it.' One patient says, 'My sister died of this at 54. There's nothing I can do.'
1. Sort these patients using a stage-of-change framework and explain why the visit format failed most of them.
Read the worked answer
The single-visit action plan presumes action, but the distribution suggests most patients are earlier:
- 'One more thing' / app never opened, no intention formed → precontemplation or early contemplation.
- 'My sister died at 54, nothing I can do' → low self-efficacy plus external/stable causal attribution; still precontemplation, with a specific barrier to address.
- The 31 who are following it → action.
- 'I already blew it' after two missed days → people who entered action and then met the abstinence violation effect.
Matching a treatment to a stage is the core point: action-oriented plans given to precontemplators typically produce no behavior change and can reduce engagement, because the plan implies a decision the patient has not yet made. The 140 unopened apps are not defiance; they are a stage mismatch. What predicts movement at precontemplation is raising awareness and eliciting the patient's own reasons, not supplying technique.
2. Explain the 'I already blew it' pattern and what it predicts about how relapses should be handled.
Read the worked answer
This is the abstinence violation effect: after a lapse, the person makes an internal, stable, global attribution ('I have no self-control'), experiences guilt, and abandons the whole effort rather than resuming. The all-or-nothing framing of the 12-week plan supplies exactly the cognitive setup for it - every day is either on-plan or a failure, with no category for a partial lapse.
Prediction: the more rigid and total the plan, the larger the collapse after the first small violation. That is measurable - patients with a lapse in week 1–2 should show disproportionate dropout.
Practical implication: build explicit lapse-recovery rules into the plan ('missing two days is expected; here is the restart step'), attribute lapses to the situation and to a modifiable technique rather than to character, and keep the criterion graded - a percentage of days logged rather than a streak. Handling the lapse is a distinct skill from adhering.
3. Design the next visit around the patient who says, 'My sister died of this at 54. There's nothing I can do.'
Read the worked answer
That statement carries a stable, external, uncontrollable causal model and low self-efficacy; arguing against it directly tends to produce reactance and entrenchment. Better sequence:
- Elicit the patient's own account and reflect it - 'you've watched this kill someone you love, and you're not convinced anything you do changes that.' Reflection rather than confrontation is the core of motivational interviewing.
- Decouple the sister's outcome from this patient's controllability by asking what differed - age at diagnosis, other conditions, access to care - without asserting optimism.
- Develop discrepancy between the stated fatalism and anything the patient does value (grandchildren, work, avoiding dialysis), using the patient's words.
- Support autonomy explicitly - 'it's your call' - which lowers reactance and makes any commitment self-attributed rather than compliance.
- Only then, if the patient signals readiness, offer a small, concrete, high-probability-of-success first step (one medication at one time of day), framed as an experiment.
Explicitly avoid: information-dumping, fear appeals, and a full 12-week plan. The ordering - evocation before advice - is the reasoning being tested, not the particular script.