🏥 Chapter 4: Clinical Assessment Tools

Standardized Depression Screening Instruments

Clinical assessment tools provide standardized, validated methods for quantifying depression severity and tracking treatment response. When digitized and integrated with AI systems, these instruments enable scalable screening, continuous monitoring, and data-driven clinical decision support. Understanding these tools is essential for implementing evidence-based depression detection systems.

PHQ-9: The Gold Standard for Digital Implementation

The Patient Health Questionnaire-9 (PHQ-9) has emerged as the preferred screening tool for digital depression detection systems due to its brevity, strong psychometric properties, and public domain status. Developed by Kroenke et al. in 2001, the PHQ-9 directly corresponds to the nine DSM-5 diagnostic criteria for major depressive disorder.

PHQ-9 Structure and Scoring

Each of the nine items asks how often over the past two weeks the respondent has been bothered by specific symptoms, with responses scored as:

• 0 = Not at all
• 1 = Several days
• 2 = More than half the days
• 3 = Nearly every day

Digital PHQ-9 Implementation

class DigitalPHQ9:
    """Digital implementation of PHQ-9 with scoring and interpretation"""
    
    QUESTIONS = [
        "Little interest or pleasure in doing things",
        "Feeling down, depressed, or hopeless",
        "Trouble falling/staying asleep, or sleeping too much",
        "Feeling tired or having little energy",
        "Poor appetite or overeating",
        "Feeling bad about yourself or that you are a failure",
        "Trouble concentrating on things",
        "Moving or speaking slowly, or being fidgety/restless",
        "Thoughts that you would be better off dead or hurting yourself"
    ]
    
    def __init__(self):
        self.responses = []
        
    def present_question(self, question_index):
        """
        Present single PHQ-9 question with response options
        Returns formatted question for display
        """
        if question_index >= len(self.QUESTIONS):
            raise ValueError("Invalid question index")
            
        return {
            'question_number': question_index + 1,
            'text': self.QUESTIONS[question_index],
            'timeframe': 'Over the last 2 weeks, how often have you been bothered by:',
            'options': [
                {'value': 0, 'label': 'Not at all'},
                {'value': 1, 'label': 'Several days'},
                {'value': 2, 'label': 'More than half the days'},
                {'value': 3, 'label': 'Nearly every day'}
            ]
        }
    
    def record_response(self, question_index, response_value):
        """Record user response with validation"""
        if not 0 <= response_value <= 3:
            raise ValueError("Response must be 0-3")
        
        self.responses.append({
            'question': question_index,
            'value': response_value,
            'timestamp': datetime.now()
        })
    
    def calculate_score(self):
        """Calculate total PHQ-9 score and severity category"""
        if len(self.responses) != 9:
            raise ValueError("All 9 questions must be answered")
        
        total_score = sum(r['value'] for r in self.responses)
        
        # Determine severity category
        if total_score < 5:
            severity = "Minimal/None"
            recommendation = "No treatment indicated"
        elif total_score < 10:
            severity = "Mild"
            recommendation = "Watchful waiting; consider psychotherapy"
        elif total_score < 15:
            severity = "Moderate"
            recommendation = "Psychotherapy or medication recommended"
        elif total_score < 20:
            severity = "Moderately Severe"
            recommendation = "Psychotherapy and medication recommended"
        else:
            severity = "Severe"
            recommendation = "Psychotherapy and medication recommended; consider intensive treatment"
        
        # Check for suicidal ideation (question 9)
        suicide_risk = self.responses[8]['value'] > 0
        
        return {
            'total_score': total_score,
            'severity': severity,
            'recommendation': recommendation,
            'suicide_risk_flagged': suicide_risk,
            'subscale_scores': self._calculate_subscales(),
            'completion_date': datetime.now()
        }
    
    def _calculate_subscales(self):
        """
        Calculate cognitive-affective and somatic subscales
        Helps identify symptom patterns
        """
        cognitive_affective = sum(self.responses[i]['value'] for i in [0, 1, 5, 8])
        somatic = sum(self.responses[i]['value'] for i in [2, 3, 4, 6, 7])
        
        return {
            'cognitive_affective': cognitive_affective,
            'somatic': somatic,
            'ca_percentage': (cognitive_affective / 12) * 100,
            'somatic_percentage': (somatic / 15) * 100
        }
        

Computerized Adaptive Testing (CAT)

Computerized Adaptive Testing represents the next evolution in digital assessment, using algorithms to select questions dynamically based on previous responses. CAT can achieve the same measurement precision as full-length assessments using 50-60% fewer items, reducing respondent burden while maintaining accuracy.

How CAT Works for Depression Assessment

• Item Response Theory (IRT) Foundation: Each question is calibrated for difficulty and discrimination based on large validation datasets
• Dynamic Question Selection: Algorithm selects next question based on current severity estimate to maximize information gain
• Precision-Based Termination: Testing stops when severity estimate reaches desired precision threshold
• Continuous Scoring: Provides granular severity scores rather than discrete categories

Ecological Momentary Assessment (EMA)

Ecological Momentary Assessment involves repeated sampling of participants' current symptoms, behaviors, and contexts in real-time, in natural environments. EMA overcomes recall bias inherent in retrospective questionnaires and captures symptom variability across time and situations.

EMA Implementation for Depression Monitoring

Sample EMA Questions

class DepressionEMA:
    """Ecological Momentary Assessment for depression monitoring"""
    
    # Brief mood questions for multiple daily assessments
    BRIEF_QUESTIONS = [
        {
            'id': 'mood_current',
            'text': 'How are you feeling right now?',
            'type': 'slider',
            'scale': [0, 10],
            'labels': ['Very sad', 'Neutral', 'Very happy']
        },
        {
            'id': 'energy_current',
            'text': 'How is your energy level?',
            'type': 'slider',
            'scale': [0, 10],
            'labels': ['Exhausted', 'Normal', 'Energetic']
        },
        {
            'id': 'activity_current',
            'text': 'What are you doing?',
            'type': 'multiple_choice',
            'options': ['Working', 'Socializing', 'Leisure', 'Resting', 'Other']
        },
        {
            'id': 'social_context',
            'text': 'Who are you with?',
            'type': 'multiple_choice',
            'options': ['Alone', 'Family', 'Friends', 'Coworkers', 'Strangers']
        }
    ]
    
    # End-of-day comprehensive questions
    EOD_QUESTIONS = [
        {
            'id': 'day_mood_avg',
            'text': 'Overall, how was your mood today?',
            'type': 'slider',
            'scale': [0, 10]
        },
        {
            'id': 'day_stress',
            'text': 'How stressful was today?',
            'type': 'slider',
            'scale': [0, 10]
        },
        {
            'id': 'social_interaction',
            'text': 'Did you have meaningful social interaction today?',
            'type': 'yes_no_scale',
            'scale': [0, 4],
            'labels': ['None', 'A little', 'Some', 'Quite a bit', 'A lot']
        },
        {
            'id': 'sleep_quality_previous',
            'text': 'How well did you sleep last night?',
            'type': 'slider',
            'scale': [0, 10],
            'labels': ['Very poorly', 'Okay', 'Very well']
        }
    ]
    
    def schedule_assessments(self, strategy='signal_contingent', n_daily=4):
        """
        Schedule EMA assessments based on selected strategy
        """
        if strategy == 'signal_contingent':
            # Random times between 9 AM and 9 PM, at least 2 hours apart
            import random
            schedule = []
            current_hour = 9
            for i in range(n_daily):
                hour = current_hour + random.randint(0, 2)
                minute = random.randint(0, 59)
                schedule.append(f"{hour:02d}:{minute:02d}")
                current_hour = hour + 2
            return schedule
            
        elif strategy == 'end_of_day':
            return ['20:00']  # 8 PM daily
            
        elif strategy == 'context_aware':
            # Triggered by detected low activity or social isolation
            return 'sensor_triggered'
        

Integration with Electronic Health Records

Integrating depression screening tools into Electronic Health Record (EHR) systems enables routine screening in primary care and specialty settings. This integration supports population health management and clinical decision support.

EHR Integration Best Practices

• Workflow Integration: Embed screening into check-in process, pre-visit tablets, or patient portals to minimize clinician burden
• Automated Scoring: Instant calculation and interpretation with severity flagging for clinician review
• Decision Support Alerts: Pop-up alerts for positive screens (PHQ-9 ≥10) or suicidal ideation (item 9 >0)
• Longitudinal Tracking: Graph scores over time to visualize treatment response and identify deterioration
• Care Coordination: Automated referrals to mental health specialists triggered by severity thresholds
• Quality Metrics: Track screening rates and follow-up for quality improvement and value-based care reporting

Suicide Risk Assessment

Item 9 of the PHQ-9 screens for suicidal ideation: "Thoughts that you would be better off dead, or of hurting yourself in some way." Any positive response (score >0) requires immediate clinical follow-up. Digital systems must have protocols for suicide risk management.

Digital Suicide Risk Protocols

• Immediate Flagging: Any positive response on suicide item triggers immediate alert
• Crisis Resources Display: Automatically present National Suicide Prevention Lifeline (988), Crisis Text Line (text HOME to 741741), and local emergency resources
• Clinician Notification: Real-time alert sent to designated clinician for same-day contact
• Secondary Assessment: Follow-up with Columbia-Suicide Severity Rating Scale (C-SSRS) for risk stratification
• Safety Planning: Digital safety plan creation with warning signs, coping strategies, support contacts
• Elevated Monitoring: Increase EMA frequency and passive monitoring sensitivity for individuals flagged at risk