.png){kind=small}
Using AI to Support Medication Adherence and Mood Tracking
- Dr Titilayo Akinsola
- Nov 2
- 6 min read
Introduction
When managing mental health, one of the most persistent challenges is staying consistent with medications and tracking subtle mood changes. Missed pills, unreported mood dips, delayed check-ins: each creates risk of relapse, reduced efficacy, or functional decline. But what if technology could help? With recent advances in artificial intelligence (AI), there’s growing potential to enhance medication adherence, monitor mood and emotional fluctuations, and support clinicians and clients between sessions. In this post we’ll explore how AI works in this space, what the evidence says, how you at Favor Mental Health can leverage it (or at least understand it), and what to watch out for.
Why This Matters — The Gap in Adherence & Mood Monitoring
Studies consistently show that medication non-adherence is a major barrier in mental‐health treatment: for example, mood disorders with inadequate medication compliance have worse outcomes, higher relapse.
Mood changes often happen in the “silent” hours — outside clinic sessions. Without continuous or at least frequent monitoring, subtle deterioration may go unnoticed.
AI promises to help bridge that gap: by supporting reminders, detecting patterns, alerting clinicians to deviations, and empowering clients with self-insight.
What the Evidence Shows — AI for Adherence & Monitoring
AI for Medication Adherence
A focused review of AI-based tools for patient support found that in randomized clinical trials AI tools improved medication adherence by between ~6.7 % to ~32.7 % compared to standard interventions.
For example, an app (“MedAdhere”) in patients with schizophrenia showed adherence rates of ~94.7% vs ~64.4% in the control group over a 12-week period using AI verification of ingestion behaviour.
Machine-learning models have been used to predict adherence (e.g., in schizophrenia) based on early data, with AUC values around 0.81–0.92.
AI for Mood Tracking & Continuous Monitoring
AI models that monitor behavioural data (smartphone use, sleep, wearable signals) show promise in detecting mood disorders or emotional deterioration. For instance, a study of “personalised vs generalised” ML models for mental-health symptom detection found personalised models consistently outperformed generic ones.
While not always specific to medication, the overlap is clear: mood tracking plus adherence tracking gives a richer picture of treatment progress, risk, and adjustment needs.
Key Takeaway
AI is not yet a silver bullet, but it is showing real potential to improve adherence, support monitoring, and offer earlier intervention. The evidence base is growing, though limitations exist (see later).
How AI Works: What Tools & Models Look Like
Here are some of the mechanisms and tools used in practice (and that you can discuss with your clients):
Reminder & ingestion verification systems: e.g., apps that remind users when to take meds, ask for confirmations (camera, video, sensor), use ML to detect missed doses or abnormal patterns.
Predictive models: early-adherence data (first 7-14 days) used by ML algorithms to flag “at-risk” for non-adherence.
Mood-tracking models: AI analyses multi-modal data—smartphone usage, typing patterns, voice/text sentiment, wearable sleep/activity metrics—to detect mood/state changes.
Dashboard/Alerts for clinicians: Aggregated data showing client adherence trends + mood data + alerts when there’s deviation (missed meds, worsening mood) enabling pre‐emptive check-in.
Client-engagement features: Chatbots, conversational agents, self‐monitoring tools; for example, helping clients log mood, receive summaries, get nudges.
Integration with therapy/medication plans: The AI tool supplements, not replaces, the clinician: it supports progression tracking, adjustment decisions, medication review.
How Favor Mental Health Can Incorporate AI (or Prepare to)
As a provider committed to effective, ethical care, here’s how Favor can integrate the AI-adherence & monitoring dimension:
Educate clients: Inform clients about how AI-enhanced apps and tools work, benefits and limitations.
Offer or recommend tools: For clients on medication management, suggest adherence-support apps (compatible with local context) and mood‐tracking tools that integrate with therapy.
Use data in follow-ups: When clients use tracking tools, review adherence/mood logs during follow-ups, look for trends, adjust treatment accordingly.
Collaborate with tech: Choose platforms with privacy, transparency and adjustability (especially given data protection/resourcing in Lagos/Nigeria).
Set up protocols: Define how and when adherence alerts trigger outreach, how mood tracking flags set up extra check-in, how you’ll adjust meds/therapy when patterns show up.
Maintain human oversight: Ensure that AI outputs inform but do not dictate care. The therapeutic relationship remains central.
Practical Steps for Clients: What You Should Do
For clients reading this (or those you want to convert into appointments), here are actionable steps they can take:
Ask your clinician: “What adherence-support app or mood-tracking tool can we use together? Could I use one that integrates with your practice?”
Choose a tool that: reminds you of medication, allows you to log mood/symptoms, gives you visualised data you can bring to sessions.
Use the tool daily: take meds as scheduled, log mood/symptoms even briefly (e.g., after waking, before bed).
Review your data weekly: What days did I miss meds? What mood shifts occurred before/after? What patterns appear?
Bring data to your appointment: “Here are my adherence logs + mood logs. I missed meds on X days; my mood dipped on those days.” This allows your clinician (like Favor) to adjust treatment proactively.
Use the adherence/mood data as conversation starters:
“I missed two doses and mood got worse on those days.”
“On days when I slept poorly, I also skipped meds and my anxiety rose.”
Ask your provider about adjustment criteria: e.g., if adherence < 80 % for 2 weeks, we’ll schedule an outreach call; if mood declines > 20% from baseline for 7 days, we’ll consider interim session.
Be aware of privacy: read terms of any app, ensure data sharing aligns with your comfort and clinician’s context.
Limitations & Ethical Considerations
Important to be transparent about what AI can’t (yet) fully do — and what clinicians and clients must keep in mind:
Many AI studies depend on limited datasets; accuracy in diverse populations (especially in African/LMIC settings) remains a question. (Cambridge University Press & Assessment)
Data privacy is critical: mood/adherence data is sensitive. Tools must comply with relevant regulations and clinician/client must agree on how data is used.
Over-reliance risk: If a client feels the app “knows everything”, they might reduce direct clinician contact — which is dangerous because AI cannot capture nuance, context, trauma, life changes.
Engagement drops: Apps lose effectiveness if clients stop using them; human follow-up remains essential.
Risks of misinterpretation: AI-flagged alerts don’t equal crisis — they require clinician review.
Equity: Some clients may not have smartphones, reliable internet, or may prefer analog methods; AI tools should augment accessible care, not replace it.
Call to Action — Partnering with Favor Mental Health for Next-Level Support
If you’re on a medication plan for anxiety, depression or another mental-health condition and struggle with consistency, or you’re engaged in therapy and feel your mood tracking is too patchy, consider a specialised session with Favor Mental Health.
Schedule a paid “Adherence + Monitoring Review” appointment. In that session we will:
Review your current medication plan, adherence history, and mood tracking habits.
Introduce (or help you select) an AI-compatible tool for adherence + mood tracking that fits your context (smartphone, usability, data sharing).
Define a protocol: how we’ll review the data together, what targets we’ll set, and how we’ll adjust treatment if adherence drops or mood shifts.
Review how mood/adherence patterns link in your life: e.g., “we’ll look at days you miss meds, mood dips, sleep disruption, and create a strategic plan.”
Provide you with a monthly “data review” schedule and clarify how we’ll interpret the tracking in your future therapy or medication adjustment sessions.
Your mind, your meds and your daily rhythm matter. By combining human care, clinical insight and smart-tracking tools, we increase your chance of sustained improvement. Let’s work together to build a system — not just hope for a result.
Closing
Medication adherence and mood tracking are two critical pillars of successful mental-health management. AI offers a powerful assist: reminding you, tracking you, alerting you and your clinician to deviations. But technology alone doesn’t change outcomes — you and your care team do. At Favor Mental Health we believe in integrating innovation with empathy, context and purpose. Let’s take your care beyond check-lists and into data-informed, life-integrated practice.
Comments