Font size
WorksheetsRKE-517 (Week-11)
Total questions: 87
Worksheet time: 44mins
How might AI analytics contribute to health information systems, based on the provided material?
By improving diagnosis accuracy and treatment pathways
By replacing all medical staff
By eliminating the need for patient records
By automating hospital construction
What is the main purpose of normalizing databases to the Third Normal Form (3NF) during Architecture & Database Design?
To ensure data integrity and eliminate redundancy
To improve user experience
To collect more data
To enhance data security
Interoperability is important when sharing health data across different systems because:
It allows seamless access and sharing, improving patient care and system efficiency.
It increases data redundancy.
It makes data collection easier.
It reduces the need for database normalization.
What percentage of the world's data volume is generated by healthcare?
10%
20%
30%
50%
What is the main challenge organizations face with healthcare data?
Lack of data collection
Difficulty transforming data into actionable improvements
High cost of data storage
Limited access to patient records
What does the "DRIP" paradox refer to in healthcare information management?
Data Rich, Information Poor
Data Rare, Information Plentiful
Data Reliable, Information Powerful
Data Rapid, Information Precise
What is the critical shift needed in healthcare information management?
From System of Insight to System of Record
From System of Record to System of Insight
From System of Record to System of Storage
From System of Insight to System of Documentation
Which of the following best describes the "Old Way" of handling missed appointments in healthcare?
Predicting missed appointments before they happen
Recording a missed appointment after the fact
Ignoring missed appointments
Scheduling appointments automatically
How does the "New Way" improve upon the "Old Way" in managing missed appointments?
By ignoring missed appointments
By recording missed appointments after they happen
By predicting missed appointments before they happen and taking preventive action
By rescheduling missed appointments automatically
Why is transitioning to a System of Insight important for healthcare organizations?
It increases data storage capacity
It enables active intelligence that anticipates needs and drives better outcomes
It reduces the number of patient records
It eliminates the need for documentation
Which government document does Indonesia's digital health vision align with?
Ministry of Health's Digital Transformation Blueprint 2024
National Education Blueprint 2024
Ministry of Finance's Economic Plan 2024
National Security Strategy 2024
What is the primary purpose of using predictive analytics in Precision Public Health according to Indonesia's digital health vision?
To forecast disease outbreaks and enable proactive intervention
To increase hospital profits
To reduce the number of healthcare workers
To eliminate the need for patient data
Which of the following is NOT mentioned as a factor analyzed in Precision Public Health for forecasting disease outbreaks?
Aggregate clinic visit data
Environmental factors
Historical trends
Genetic engineering
How does Operational Optimization use AI in Puskesmas facilities?
By optimizing staff scheduling, medication inventory, and patient flow
By replacing all medical staff with robots
By eliminating the need for patient records
By increasing the number of hospital beds
Why is AI implementation now considered mandatory in Indonesia's national health resilience strategy?
It is required for developing new competencies in analytics and algorithmic governance
It is optional for hospitals to use AI
It is only for private clinics
It is used solely for marketing purposes
Suppose a Puskesmas facility wants to reduce patient wait times. Based on the digital health vision, which AI-driven approach should they prioritize?
Predictive demand modeling for patient flow optimization
Manual scheduling of appointments
Increasing the number of waiting rooms
Reducing the number of staff
Which of the following is a key characteristic of OLTP (Online Transaction Processing) systems?
Denormalized, read-optimized structure
Highly normalized database structure (3NF)
Pre-aggregated metrics for fast pattern recognition
Optimized for complex queries across large datasets
What is the primary purpose of OLAP (Online Analytical Processing) systems?
Fast INSERT, UPDATE, DELETE operations
Data integrity and consistency
Rapid access to historical patterns and aggregate statistics
Minimizing data redundancy through multiple related tables
Which of the following is a critical weakness of OLTP systems when used for AI applications?
Lack of historical data warehousing
Too many table joins slow analysis and transaction locks prevent simultaneous analytical queries
Poor data integrity and consistency
Inability to optimize for complex queries across large datasets
Why is it problematic to run AI directly on operational OLTP databases?
It improves system stability
It creates performance bottlenecks and system instability
It enhances data redundancy
It optimizes for fast pattern recognition
Compare the database structure of OLTP and OLAP systems. How does their structure support their respective purposes?
OLTP uses denormalized structure for fast queries; OLAP uses normalized structure for data integrity
OLTP uses highly normalized structure for data integrity; OLAP uses denormalized structure for read optimization
Both OLTP and OLAP use normalized structures for consistency
OLTP and OLAP both use denormalized structures for pattern recognition
Which feature is optimized in OLAP systems to support machine learning algorithms?
Fast INSERT, UPDATE, DELETE operations
Data redundancy minimization
Complex queries across large datasets
Transaction locks for simultaneous queries
Which of the following best describes the main purpose of the ETL pipeline in a clinical data environment?
To ensure AI models receive properly formatted, cleansed data without disrupting clinical operations.
To store patient records in transactional databases for future reference.
To visualize patient demographics for clinical staff.
To automate appointment scheduling for patients.
What is the primary function of the "Extract" stage in the ETL pipeline?
Pull raw operational data from transactional databases on scheduled intervals or triggered events.
Cleanse inconsistencies and calculate derived features.
Store processed data in Analytical Base Tables.
Visualize historical patterns in patient data.
During the "Transform" stage of the ETL pipeline, which of the following activities is performed?
Cleansing inconsistencies and calculating derived features such as "Lead Time" between booking and appointment.
Storing data in production systems.
Scheduling patient appointments.
Extracting raw data from external sources.
Which stage of the ETL pipeline involves storing processed data in Analytical Base Tables optimized for machine learning model training and prediction?
Extract
Transform
Load
Analyze
Why is architectural separation important in the ETL pipeline for clinical environments?
It allows clinical staff to continue using the operational system while data scientists train and deploy AI models on the analytical platform.
It enables faster appointment scheduling for patients.
It reduces the need for data cleansing in the pipeline.
It eliminates the need for transactional databases.
Suppose you are designing an ETL pipeline for a hospital. How would you ensure that historical patterns are aggregated during the transformation stage?
By applying business rules and aggregating historical data during the transformation process.
By storing all raw data in Analytical Base Tables.
By extracting only the most recent patient records.
By visualizing data using dashboards.
Which of the following best describes the concept of "Garbage In, Garbage Out" in the context of AI?
The quality of an algorithm's output is limited by the quality of the data it learns from.
AI algorithms can always compensate for poor data with sophisticated modeling.
The more data you have, the less important its quality becomes.
AI systems do not require data preparation to function effectively.
What is the main risk of missing data in patient records, according to the concept of completeness?
It increases the speed of data processing.
It creates blind spots and makes predictions unreliable for a significant portion of the population.
It improves the accuracy of SMS-based interventions.
It reduces the need for data validation.
Why is standardization important for data consistency in health information systems?
It allows models to treat identical concepts as different categories.
It fragments statistical power.
It ensures that identical concepts are treated the same, improving statistical power.
It encourages the use of multiple coding systems for the same concept.
Suppose a health dataset uses three different codes for gender: "Pria/Wanita", "M/F", and "L/P". What is a likely consequence if these are not standardized?
The model will ignore gender information.
The model will treat the same gender as different categories, reducing statistical power.
The model will automatically merge all codes correctly.
The model will increase its predictive accuracy.
A hospital finds that 40% of its patient records lack mobile numbers. What impact could this have on an SMS-based intervention prediction model?
The model will be more reliable.
The model will be unreliable for nearly half the population.
The model will not be affected.
The model will automatically fill in the missing numbers.
Which step in the data readiness process is most directly responsible for removing errors and inconsistencies from raw data?
Data Intake
Cleansing
Validation
High Quality Data
Why is achieving both high accuracy and consistency important in healthcare analytics?
It ensures that data is both correct and standardized, leading to reliable model predictions.
It allows for more creative data coding.
It reduces the need for data validation.
It makes data preparation unnecessary.
Which dimension of data readiness focuses on the need for historical context rather than just current snapshots in AI models?
Outcome Granularity
Temporality
Data Privacy
Data Volume
Why is it important for AI models to use the trend of BMI measurements over several years rather than just today's weight reading?
It helps reduce data storage costs.
It provides historical context for better prediction of future health risks.
It simplifies the data analysis process.
It eliminates the need for patient records.
What is a design implication for maintaining temporality in data for AI models?
Overwrite previous values with updates.
Use slowly changing dimensions to preserve historical states.
Delete old data regularly.
Only use current data snapshots.
What is the common trap when handling outcome granularity in appointment data for AI models?
Adding extra status flags
Hard-deleting "No-Show" appointments from the database
Using only completed appointments
Overwriting appointment times
What is the correct approach to preserve outcome granularity in appointment data for AI models?
Delete all cancelled appointments
Use status flags to preserve the ground truth and maintain audit trails
Only keep confirmed appointments
Ignore appointment status
How does preserving status flags such as 'No-Show', 'Cancelled', and 'Completed' benefit AI model training?
It reduces the amount of data to process.
It ensures only successful appointments are analyzed.
It maintains the ground truth for accurate predictions and audit trails.
It allows for faster database queries.
Which type of analytics focuses on summarizing historical performance metrics, such as monthly admission volumes or average length of stay?
Descriptive Analytics
Diagnostic Analytics
Predictive Analytics
Prescriptive Analytics
What is the main question addressed by Diagnostic Analytics?
What happened?
Why did it happen?
What will happen?
How can we make it happen?
Which analytics type uses machine learning models to forecast future outcomes with probability scores?
Descriptive Analytics
Diagnostic Analytics
Predictive Analytics
Prescriptive Analytics
Prescriptive Analytics is best described as:
Summarizing historical data
Analyzing root causes of trends
Forecasting future outcomes
Recommending specific actions to achieve desired outcomes
An organization wants to understand the drivers behind observed trends and anomalies in their healthcare data. Which analytics approach should they use?
Descriptive Analytics
Diagnostic Analytics
Predictive Analytics
Prescriptive Analytics
If a hospital wants to optimize staff scheduling based on predicted patient admission rates, which analytics level should they advance to for greatest strategic value?
Descriptive Analytics
Diagnostic Analytics
Predictive Analytics
Prescriptive Analytics
Predictive Analytics can be used to improve patient care in a hospital setting. Which of the following is a specific example of its application?
By summarizing past admission volumes
By identifying the root cause of anomalies
By forecasting readmission risk to enable early intervention
By recommending staff schedules
What is the main business problem described in the case study "The No-Show Predictor"?
High patient satisfaction rates
Persistent 20% no-show rate causing operational failures
Overbooked appointment slots
Increased staff productivity
Which of the following is a proposed AI solution to address the no-show problem?
Increase appointment reminders
Deploy a predictive model to calculate probability of patient attendance
Hire more clinical staff
Reduce appointment slots
According to the feature engineering section, what does "Lead Time" refer to?
The time a patient spends with a doctor
The days between booking and appointment date
The number of appointments a patient has missed
The time taken to confirm an appointment
Why is patient history important in predicting future no-shows?
It shows the patient's favorite doctor
Past behavior is the strongest predictor of future behavior
It reveals the patient's medical condition
It determines the patient's payment method
Which contextual factors are considered in the predictive model for patient attendance?
Patient's age and gender
Weather forecasts, public holiday calendars, and local transport disruptions
Doctor's experience level
Hospital location
(DoK Level 2) If a clinic wants to reduce its no-show rate, which feature should it consider adjusting based on the information provided?
Lead time between booking and appointment date
The color of the waiting room
The type of medical equipment used
The number of magazines in the lobby
(DoK Level 2) How might integrating contextual factors improve the accuracy of the no-show predictor model?
By making the model more complex without adding value
By considering external influences like weather and transport disruptions that affect patient mobility
By ignoring patient preferences
By focusing only on appointment times
(DoK Level 3) Imagine you are designing a predictive model for patient attendance. Which combination of features would you prioritize and why?
Lead time, patient history, and contextual factors, because they collectively address timing, behavioral, and external influences
Only patient history, because it is the easiest to collect
Only lead time, because it is the most relevant
Contextual factors alone, because they are unpredictable
(DoK Level 3) If a predictive model identifies a high probability of patient no-show, what proactive intervention strategy could a clinic implement before the appointment date?
Ignore the prediction and proceed as usual
Contact the patient to confirm or reschedule the appointment
Cancel all future appointments for the patient
Increase the appointment fee
Which action is recommended for patients classified as High Risk (>60%) in the "Smart Schedule" Traffic Light System?
Automated SMS reminder 24 hours prior
Standard email reminder only
Personal phone call from scheduling staff 48 hours before appointment
No reminder required
What is the probability range for Medium Risk in the "Smart Schedule" Traffic Light System?
Less than 30%
30-60%
Greater than 60%
10-40%
Which action is taken for patients classified as Low Risk (<30%)?
Personal phone call 48 hours before appointment
Automated SMS reminder 24 hours prior
Standard email reminder only
No action required
Why does explainability matter in the "Smart Schedule" Traffic Light System?
It allows staff to ignore risk scores
It builds user trust and enables override when clinical judgment differs
It reduces the need for reminders
It increases the number of high-risk patients
Which of the following is NOT mentioned as a factor in the tooltip for risk explainability?
Heavy rain predicted
Number of prior no-shows in last 6 months
14-day lead time
Patient's age
If a patient is assigned Medium Risk, what is the recommended action according to the system?
No reminder is sent
Standard email reminder only
Automated SMS reminder 24 hours prior with easy reply options
Personal phone call 48 hours before appointment
How does the "Smart Schedule" Traffic Light System help busy clinic staff?
By providing raw probability scores only
By giving clear, immediate guidance on which patients require attention
By eliminating the need for scheduling
By assigning all patients to high risk
What is a common issue with type 2 diabetes diagnosis according to the case study?
It often remains undiagnosed for years, leading to preventable complications.
It is always diagnosed early due to obvious symptoms.
It is only found in children.
It can be detected without any screening protocols.
Which of the following is NOT a feature used by the automated risk scoring model for diabetes screening?
Blood type
Age
Body Mass Index (BMI)
Hypertension History
How does the AI intervention help clinicians in diabetes screening?
It analyzes existing EMR data without adding data collection burden.
It requires clinicians to manually enter all patient data.
It only works for patients under 45 years old.
It delays the screening process.
Which step occurs when the risk score exceeds the threshold in the workflow integration?
An alert appears on the nurse triage screen before the physician encounter begins.
The patient is sent home without further testing.
The doctor ignores the risk score.
The nurse waits until after the physician encounter to act.
Why is automated diabetes risk scoring beneficial in clinical settings? (DoK Level 2)
It helps identify high-risk patients who might otherwise be missed by standard screening protocols.
It replaces the need for all medical staff.
It eliminates the need for any patient data.
It only works for patients with no family history of diabetes.
The integration of automated risk scoring into the clinical workflow can reduce missed opportunities for diabetes screening by:
Providing real-time alerts before the physician encounter, allowing nurses to order necessary tests early and ensuring timely diagnosis.
Delaying the alert until after the patient leaves, ensuring no unnecessary tests are ordered.
Only alerting the physician after all appointments are completed.
Requiring manual review of all patient records before each visit.
What is the main reason physicians prefer narrative documentation over structured ICD-10 codes?
It captures clinical nuance and reasoning.
It is faster to complete.
It is required by government regulations.
It eliminates administrative burden.
Which step in the NLP solution process involves identifying key symptoms and anatomical references?
NLP Analysis
Physician Input
AI Output
Physician Action
What does the AI output suggest in the example process flow, and with what confidence level?
"Angina Pectoris (I20.9)" with 87% confidence
"Diabetes Mellitus (E11)" with 95% confidence
"Hypertension (I10)" with 80% confidence
"Asthma (J45)" with 90% confidence
Why is human verification critical in the medical coding process using AI?
To ensure legal and clinical responsibility remains with the clinician
To speed up the coding process
To allow auto-coding without errors
To reduce the need for patient care continuity
Natural Language Processing (NLP) bridges the gap between clinical expression and administrative requirements in medical coding by:
Analyzing free-text clinical notes to extract diagnostic concepts and suggest appropriate ICD-10 codes
Automatically coding all patient records without review
Replacing physicians in the documentation process
Ignoring clinical nuance and focusing only on billing codes
Which of the following best describes algorithmic bias in AI systems as discussed in the material?
AI systems can amplify existing disparities if they learn from biased historical data.
AI systems always reduce disparities in healthcare.
Algorithmic bias only occurs in financial applications, not healthcare.
AI systems ignore historical data when making decisions.
Why are fairness audits across demographic groups considered mandatory before AI deployment, according to the material?
To ensure the AI is faster than humans
To prevent the AI from learning incorrect correlations and causing unfair outcomes
To increase the number of appointments
To reduce the cost of healthcare
If you were tasked with deploying an AI system in healthcare, what strategic step would you take to avoid discriminatory outcomes based on the information provided?
Ignore historical data and focus only on current patient behavior.
Conduct fairness audits across demographic groups before deployment.
Double-book appointments for all districts.
Assign risk scores randomly to patients.
What is automation bias?
The tendency to favor suggestions from automated systems, even when they contradict independent judgment or observation.
The tendency to ignore automated systems in favor of human judgment.
The tendency to always question automated systems.
The tendency to rely only on manual procedures.
Which of the following best describes a risk of automation bias in healthcare?
Improved patient safety due to faster decisions.
Significant patient safety risks due to over-reliance on automated suggestions.
Reduced need for professional judgment.
Elimination of all diagnostic errors.
In the example scenario, what was the consequence of the physician trusting the AI diagnostic support system?
The patient's condition improved.
The physician ordered unnecessary tests.
The patient's condition worsened.
The physician ignored the algorithm completely.
According to Medical Practice Law, who bears full legal responsibility for clinical outcomes?
The AI system
The hospital administrator
The physician
The patient
What is the role of AI systems classified as Clinical Decision Support Systems (CDSS) in medical practice?
They replace professional judgment.
They inform decisions but never replace professional judgment.
They make final decisions for physicians.
They are used only for administrative tasks.
Under Indonesia's Personal Data Protection Law, what must patients be informed about regarding AI?
The cost of AI systems.
When AI contributes to their care decisions.
The technical details of the AI algorithm.
The manufacturer of the AI system.
What right do patients have under Indonesia's Personal Data Protection Law regarding AI involvement?
The right to ignore AI recommendations.
The right to request human-only review.
The right to modify the AI algorithm.
The right to choose their physician.
Why must consent forms explicitly mention algorithmic involvement under UU PDP Transparency Requirements?
To ensure patients are aware of AI's role in their care decisions.
To reduce paperwork for physicians.
To allow AI systems to operate without oversight.
To speed up the consent process.
Which of the following is a requirement during Pre-Deployment Validation in the Ethical Governance Framework for Health AI?
Ensure accuracy disparities don't exceed 5% between populations.
Track real-world performance metrics monthly.
Include patients in governance committees.
Require every prediction to have three contributing factors.
