Clinician Guide

What Your Patients Are Tracking

Assessment Instruments

All scoring algorithms follow published clinical manuals exactly. No machine learning or AI is used — all calculations are deterministic and match the published instruments. Cutoff scores and severity bands are not modified or approximated. Screening disclaimers display before and after every assessment.

PHQ-9 and GAD-7 were developed by Drs. Robert L. Spitzer, Janet B.W. Williams, and Kurt Kroenke, with support from Pfizer Inc. STIPO-R is used under guidelines from the International Society of Transference-Focused Psychotherapy (ISTFP).

On iOS 18 and later, patients may opt to write PHQ-9 and GAD-7 results back to Apple Health as HKPHQ9Assessment and HKGAD7Assessment scored assessment types, along with State of Mind entries from mood logs.

Using the Data in Sessions

Mood Trends

Patients can show you their Insights Dashboard, which visualizes mood patterns over time. The circumplex view plots valence (pleasant–unpleasant) against arousal (activated–deactivated), helping identify emotional patterns that narrative recall alone might miss. A Year in Pixels view provides a full-year color-coded mood grid for longitudinal context.

Assessment Tracking

Longitudinal assessment scores are charted automatically. Patients taking a PHQ-9 biweekly can show you three months of trend data in seconds — far more useful than a single-session score.

Correlation Data

The app surfaces statistical correlations between mood and other factors (sleep, exercise, medication adherence, HealthKit metrics, blood pressure variability, sleep architecture, HRV, and more). Key safeguards:

• Minimum 7 data points per group before any insight appears
• Only correlations with |r| ≥ 0.3 (moderate or stronger) are shown
• Language always says "associated with" — never "causes"
• Sample sizes are displayed alongside every correlation

Medication Vital Effects

For patients on medication, the app performs time-windowed pre/post dose analysis (paired t-test) to identify statistically detectable effects on heart rate, HRV, blood pressure, sleep, stress, and body activation. Particularly useful for cardiovascular and psychiatric medications.

Daily Recovery Score & Stress

Patients who wear an Apple Watch can track a Daily Recovery Score (0–100) computed from overnight vitals. They can also capture perceived felt stress on 0–10 sliders to identify divergence between physiological and subjective stress states ("Mind-Body Match") — valuable for patients with alexithymia or interoceptive difficulties.

Defense Mechanism Tracking

For clinicians working within a psychodynamic framework, patients can log defense mechanisms they notice between sessions. This accelerates the development of the observing ego — the ability to notice one's own psychological processes.

Psychodynamic Writing Prompts

The journal includes rotating prompts across six psychodynamic categories: Free Association, Inner World, Relational, Affective, Self-Reflection, and Symbolic. Prompts adapt to the patient's age group and emotional context.

Sentence Completion Exercise

An RISB-inspired projective assessment with 40 original stems across four domains (Family/Attachment, Social/Interpersonal, Self-Concept, General Attitudes). Uses on-device Natural Language processing (Apple's NLTagger) for sentiment analysis and keyword extraction across 11 themes, producing a deterministic adjustment score (0–100). Particularly useful for patients who have difficulty articulating affect directly.

Body Sensation Map

The interactive body map helps patients with high somatic anxiety identify patterns between physical sensations and emotional states. Interoceptive awareness data collected over time can reveal patterns not apparent in session.

Clinical Formulation Report

Premium users can generate an eleven-section structured psychodynamic summary drawn from their own existing data. Sections include Presenting Picture, Affect Regulation, Defensive Functioning, Personality Organization, Mentalizing Capacity, Dream Content, Behavioral Correlates, Longitudinal Trajectory, Clinical Impressions, and Data Completeness. The report is generated on-device using deterministic rule-based logic — no machine learning or AI interpretation is involved.

Data Export for Clinical Use

Patients can export their data in two formats from Settings → Data Management → Export Data:

• PDF report — a human-readable summary suitable for clinical records, including mood trends, assessment scores, correlations, and optional Clinical Formulation sections
• JSON export — machine-readable format for research or integration with other clinical tools

Exports are generated on-device. No data passes through any server during the export process.

Sample “Show Your Doc” Report

This is what your patients can generate and bring to sessions. The report combines mood data, assessment scores, correlations, and psychodynamic formulation sections into a single clinical summary.

Safety Features

The app includes automatic safety check flows triggered by specific clinical thresholds. You should be aware of these so you understand what your patient experienced between sessions.

When triggered, the app displays a full-screen safety check (cannot be dismissed by swiping) with crisis resources including the 988 Suicide & Crisis Lifeline and the Crisis Text Line (text HOME to 741741). The app records only a structured audit event with three fields (type, source, outcome) — it never stores the patient's specific emotion labels, assessment responses, or verbatim text.

Safety features are always free, cannot be disabled, and cannot be paywalled. This is architectural, not policy.

Privacy Architecture

Understanding the app's privacy model helps you recommend it with confidence:

• All data stays on-device. There are no servers that receive or process patient data.
• iCloud sync is optional and uses end-to-end encryption via Apple CloudKit.
• Stress and recovery data is stored in a separate local-only SwiftData store that is never synced to CloudKit.
• No analytics SDKs. No Google Analytics, Firebase, Mixpanel, Sentry, Crashlytics, or third-party trackers.
• No account required. No email, no username, no identifying information collected.
• HealthKit data is read-only (with optional iOS 18+ writes for State of Mind and PHQ-9/GAD-7 assessments) and never transmitted externally.
• No machine learning or AI is used for clinical decisions. All scoring and safety logic is deterministic and rule-based.
• We cannot access patient data under any circumstances. This is architectural, not policy.

Full details are in our Privacy Policy.

Age-Adaptive Interface

The app adapts its interface, vocabulary, and available features based on the user's age group:

Age is determined by date of birth entered at onboarding. Parental consent is required for users under 13 (COPPA compliance), and the Texas SCOPE Act (HB 18) provisions apply to all users under 18: no targeted advertising, no profiling, no commercial data sharing, no dark patterns.

Recommending to Patients

When recommending The Observing Ego to patients, you may find these talking points helpful:

• "It's a mood tracker that keeps everything on your phone — no accounts, no data sharing."
• "You can track your mood, habits, and medications, then bring the data to our sessions."
• "It includes clinical screening tools like the PHQ-9 and GAD-7 that we can track over time together."
• "If you ever log something that suggests you might need immediate support, the app will show you crisis resources automatically — and those resources are always free."
• "There's a free tier that covers the basics. Premium unlocks longitudinal insights and the clinical formulation report."

Questions or Feedback

We welcome input from clinicians. If you have questions about the app's clinical content, suggestions for improvement, or want to discuss integration with your practice: